Final Report Summary - SLOPE (Integrated proceSsing and controL systems fOr sustainable forest Production in mountain arEas)
Executive Summary:
Forest production in mountains has a significant role in the world economy, being these areas 23% of the total Earth forest. Projections suggest that this trend will not change before 2030 with the major production growth in Europe, Russia, Eastern Europe and South America. This growth involves all the major wood sectors like wood-based panels, paper, industrial round wood and bioenergy and is a consequence of increasing energy demands interest on wood properties and in the construction sector. However, wood production in slope conditions differs from traditional one for the limits imposed by terrain morphology, for complex operation planning and hardware requirements, like the use of manual chainsaws and cable cranes instead of processor heads adopted in flatlands. Powerful and intelligent machines must be developed for forest operations in steep terrain together with improved types of workflows.
The SLOPE project fills this gap by developing a new process and an integrated information system for the optimization of forest production where remote sensing (UAV and Satellites Imageries) and on-field surveying systems (Terrestrial Laser Scans) are used to build a detailed 3D forest model to support the entire wood process, from tree RFID tagging to cut and debranched logs sent to the sawmills.
The adoption of this model, enables forest data sharing among the involved actors through mobile devices and complex harvesting machines, improves wood output, adapting the available wood quality with the market demands, reduces manual work, increases productivity and speeds up monitoring and quality assessment thanks to detailed interactive and always accessible planning tools.
Enhanced machines based on market available devices have been built to analyse and track in real-time trees and logs. An intelligent processor head, equipped with state of the art sensors identifies the marked trees suggesting how to optimally cut it into logs and evaluates in real-time their quality during cutting and delimbing phases, combining the index quality result coming from different measurement technologies. Smart RFID tags track trees from the forest to the sawmills, with readers mounted on foresters’ roughed tablets, cable crane systems and smart-trucks, while an enhanced self-motorized carriage equipped with new remotely controlled and programmable automatic chockers increases the operational security.
Long-term planning as well as auction and selling optimization has been achieved through the connection of the digital forest model with state of the art forest planning systems integrated inside a wood trading portal, launched on the online market during the course of the project as a first example of the projects’ exploitable foreground.
Extensive trials have been performed on two test fields under different environmental conditions to demonstrate the effectiveness of the entire system and improve its reliability, opening the door to future results exploitation.
Project Context and Objectives:
Mountains occupy about 35% of the European territory and are mostly covered by forests. Forestry operations in mountain areas are seldom performed by fully mechanised systems typical of the industrial sector. The majority of operations is still characterized by manual felling and extraction of timber by cable cranes. In fact, due to the limits posed by steep terrain, non-uniform vegetation, long vegetative succession, poor road networks, hydro-geological constraints and limited storage areas on the field, harvesting and extraction are more expensive and less flexible compared to the cut-to-length systems based on wheeled machines, commonly used in flatland forests of European Nordic countries.
In order to be competitive, powerful and intelligent machines must be developed for forest work in steep terrain. The SLOPE project integrates information from remote sensing and on-field surveying systems, to support analysis for the characterization of forest resources. Spatial information can be integrated with multi-sensor data in a model for sustainable forest management and for optimization of logistics during forest operations.
The mission of SLOPE is therefore to develop an integrated system that increases the efficiency of harvesting and logistic operations and optimizes timber production in mountain areas reinforcing the communication between its actors and filling the existing gap between more expensive, less flexible forestry operations in mountain areas and optimized cut-to-length systems, commonly found in flatland forests. For this purpose, spatial information and multi-sensor data from remote sensing, UAVs (Unmanned Aerial Vehicles) and TLS (terrestrial laser scanner) surveying systems will be integrated in a digital forest model at the service of planning, monitoring and optimization of forest operations.
SLOPE focuses on adding value to the mountain forest production through reliable stumpage price evaluation and quality control of the harvested material, thanks to the integrated use of real-time novel sensing technologies inside forest machineries to improve log/biomass segregation, have up-to-date forest inventories and refine stand growth and yield models. To achieve this, intelligent sensor systems will be integrated in the cable crane/processor head/truck machines to acquire wood properties and build quality prediction models. In addition, tracking systems will be used to trace the material, from the site throughout the whole supply chain.
The collection and integration of forest information like topography, geomorphology, cadastre, roads, climate, etc. with harvesting data like extraction distance and direction, stand density, silvicultural management and with material origin, quality and availability, in a unique online interactive system, is another priority of the project. A detailed forest model containing this information will be accessible and available in real-time to a number of operators, like: logistic operators, brokers, forest owners and sawmills to be used as simulation, short/long term planning and monitoring tool for the whole production chain.
The creation of this innovative workflow requires the achievement of a number of objectives, starting from the delivery of an integrated forest surveying system to cope with the frequent lack of precise and up-to-date information about a specific stand. The formalization of a model-based forest operational planning system is then fundamental for the correct exploitation of the forest stand data in tasks such as simulation and planning of manual and machines operations on the field. Another step is the integration of novel intelligent components in harvesting machines with a focus on the delivery of a multi-sensor model-based quality control system for mountain forest production. Tree and log quality, as well as machine operations should then be monitored through a specific and advanced enterprise resource planning (ERP) system. Integration and validation of the developed systems in real-life scenarios constitutes the final step to be achieved.
The typical scenario in which SLOPE tries to adds its innovations starts from the forest property. The owner who desires to manage the forest for productive use, needs to define a plan to efficiently and correctly exploit his property. These can be “forest management" for long term planning and “pre-operational”, created before the harvesting activity begins with the purpose to optimize the production benefit, selecting the trees that meet the market requirements. The harvesting plan is done by professional foresters (private or public) which generally will perform at least a basic inventory study, an analysis of age, species, terrain characteristics, natural boundaries, environmental characteristics, soil type and health, forest and public road network as well as any other relevant facility. Harvesting plans are generally related to single parcel properties where the harvesting operation will take place.
Once defined, the parcels close to the harvesting time are visited by a forester who examines the plot for marking the trees for a more detailed plants inventory. This operation will generate a more accurate estimation of the commercial value of the standing trees to be harvested, the so-called stumpage value. Tree marking is probably the most important silvicultural task, since the identification of which tree can be harvested and which is supposed to remain, allows to define the future development of the forest and its characteristics (economic, environment and soil protection, etc.) but will also influence the following steps of the harvesting process.
Current forest marking is a manual process performed using models based on tree diameter and height to estimate the timber volume and timber products. These models have a limited accuracy due to the fact that are created based on regional information and they might be created years before the harvesting activities. Additionally, visual evaluation of standing trees to be harvested, even if done by trained and expert operators is still a subjective task which can be prone to errors and biases. Current digital measurement techniques such as Terrestrial Laser Scanner (TLS), UAV and 3D modelling can increase the accuracy of the pre-harvest timber product estimations. The right estimation of the timber assets is very important for forest owner’s incomes, as usually based on the result of the inventory, using the stumpage value as a reference, the standing forest is sold in an auction. This might be missing, for instance when the forest owner is also undertaking harvesting operations.
Harvesting activities and log storage are usually performed by specialized companies entering the parcel for planning operations and asking for cutting permissions to the competent authority. This is the stage when the cable yarder lines and their hauling corridors are designed, together with landing area, storage pads and other logistic aspects like trucks or tractors turning points, more suitable machines, etc. All of this planning is done in person, walking in the forest and with the support of maps and compasses. GPS devices are becoming the standard for identifying the boundaries of the parcel and move throughout the forest to plan the cable lines, but they are subject to errors and wrong evaluations of working areas, can lead to significant delays, to cope with unforeseen situations during the actual harvesting days.
The harvesting operations begin when all the above tasks are successfully accomplished and the season/weather conditions allow safe operations without snow or heavy rain. Manual felling of marked trees is followed by their transport to the landing area with cable carriage, where the forest workers are instructed for the cut-to-length according to production plans and customer expectations, taking in consideration the individual properties and the presence of defects inside each log. The harvesting products include timber/logs and residues, which can be converted in fuel biomass. Both products must be promptly removed due to the limited storage area typically available at landing pads in mountain forests. A proper cut of each tree is fundamental for the maximization of logs value and meet the market demands but is not always performed in the correct way or taking in account tree characteristics (e.g. branchiness, diameter, stem taper, etc.).
Usually, residues are directly chipped and delivered to the end user. The log assortments are piled in intermediate storage areas, possibly close to public roads and accessible to timber-trucks. Volume, quality classes and overall value of the piled timber is calculated at the intermediate storage area by a professional forester, providing a detailed description of the assortments and their corresponding value and the stock may be then sold in further auctions. Even in this case, the evaluation relies on the experience of the forester and can be biased. If a sawmill has not directly bought the standing forest and contracted a forest company for the harvesting operations, the end user will contract a company for transportation to the facility deposit yard once purchased the desired assortments and volume. Here the logs are further sorted according to the requirements and the organization of the yard. This raw material will then be available as an operative stock for processing according to the industrial requirements of the end user. The stocking of logs on temporary storage areas can leave them under bad weather conditions that might affect their quality and value (e.g. parasites attacks, fungi, etc.) and must be reduced to the minimum.
The described workflow highlights a number of weaknesses impacting the forest production process addressed by the SLOPE main objective: to improve, harmonize and optimize the wood processing chain from the forest to the market. This can be achieved building an integrated process which fosters actor’s interaction and exchange of information, an integrated control system for real-time quality assessment and tracking, promoting sustainable forest production through transport and market optimization with a particular focus on mountain areas competitivity.
Project Results:
The SLOPE project studied and designed during the course of three years different technical and scientific solutions for the limitations affecting forest production in mountain areas. The key results can be summarized with a reference to the overall objectives of the project.
The first one consisted in the delivery of an integrated forest surveying system for mountain areas. SLOPE successfully investigated and tested on the field the combination of traditional surveys (forest plans, satellite imageries) with new sources of information (UAVs and Terrestrial Laser Scanning surveys). The acquired information combined to provide a multi-scale accurate description of the spatial and physical characteristics of the forest stand, resulted in a digital forest model with single tree level accuracy for quality (diameter at breast height, height, branchiness) and cutting instructions (distances between cuts for logs production).
The second objective, focused on the formalization of a model-based forest operational planning system for mountain areas. In this context, a final version of the SLOPE 3D harvesting planning tool has been developed and integrated inside the SLOPE portal. This last release introduced features like: rope launcher interface and cableway interactive placement configuration, visualization of detailed stem data and short-time cost estimation (felling, cableway setup, etc.), tree virtual marking, stand spatial queries, statistics and points of interest visualization, harvesting monitoring, routing and cost transport simulation, truck movements detection and point cloud visualization. The main innovation resides in the possibility to directly plan the operation on an online 3D graphical environment, simplifying the management of these complex operations especially on the mountainous environment where the third dimension becomes essential to correctly plan the activities.
The integration of novel intelligent systems in harvesting machines operating in mountain areas has been achieved through the creation or enhancement of ready to the market machines with features tailored to automatic log quality grading and traceability, like:
• Synthetic rope launcher to speed up the deployment of the main cable carriage line
• Intelligent cable carriage system with RFID antenna for log traceability, embedded computer for data recording and a Wi-Fi antenna for processor head communication
• Automatic chockers connected with the cable carriage for safe unlocking of logs after transport
• Enhanced processor head equipped with advanced sensors for automatic quality grading and RFID tagging
• Intelligent tracking system for logs transport based on an integrate board computer to be mounted on trucks
The aforementioned machines have been subject to scientific researches for the delivery of a multi-sensor model-based quality control system. More in details, SLOPE partners have defined a series of methodologies and indexes that can support the definition of the final quality-control for the wood production on the mountain areas. Four indices have been tested, based on NIR spectroscopy, hyperspectral imaging, stress-wave measurements and cutting forces. Each one calibrated to provide an output between 0 (bad quality) and 1 (excellent quality) for each processed log. The aggregation of these indices can provide an overall log-grading index. First promising results have been presented although further trials are needed to better calibrate the model. The use of robust RFID tags and readers across the whole production, certificates the origin and quality of each log to the final customer.
Another important aspect considered within the project has been the customization of existing enterprise resource planning (ERP) systems for mountain forest. In this case, the woodland manager platform has been improved with the integration of the 3D harvesting planning tool and the long-term planning IPTIM cloud service. The outcome, released on the market as Wuudis online marketplace, has been a comprehensive web portal with three main sections: marketplace, harvesting planner, and optimizer.
For the integration and validation in real-life scenarios, the most important project goal, the SLOPE consortium deployed the whole engineered system in two test fields under very different environmental conditions: summer in Italy and winter in Austria. The first one was fundamental to raise a number of system weaknesses that the consortium was able to solve for the second one. Overall, two weeks of testing were made, processing 27 trees and about 20 logs. The whole sensoring system collected data from eight logs, also used for comparing quality assessment with standard manual procedures. The rope launcher easer has been tested for one launch and the iTruck system was tested simulating two roundtrips from the forest to the sawmill. Cable carriage worked on the two pilots with only one day of interruption. Full wood traceability was achieved and preliminary test on automatic grading showed promising results although more tests on the field would have been advisable.
At the early stage of the SLOPE project, the consortium focused the identification of target users, understanding their requirements among the entire wood processing workflow and the prerequisites for the definition of the first integrated SLOPE system like the end-user interface, the system architecture, the common data model and the hardware to be installed/adapted/deployed during the project.
On a first instance the consortium identified the set of actors/stakeholders involved within the project and, after a description of the entire SLOPE scenario, defined specific questions targeted to each different kind of end-user. The answers were collected, aggregated and summarized in a set of guidelines to be taken into account during the course of the project.
After that, an analysis of the available hardware and software brought to the definition of the hardware required for the implementation of the SLOPE optimizations among the wood processing workflow going from the specification of existing hardware like UAV platforms or Terrestrial Laser Scanning systems to new hardware combinations like those adopted for the enhanced processors head or the solution for RFID tracking.
More in details, the sensors data transfer protocols and the compatibility with the control system of the processor head, lead the consortium to identify the CompactRIO of National Instruments as the development platform for both control system and sensors data acquisition and analysis, while the RFID system research on tree marking tags, field readers, mobile devices and hardware for intelligent trucks, found an appropriate and versatile alternative using Raspberry. A toolkit to develop devices prototypes integrating different signals and communication protocols for RFID reader, Bluetooth dongles, GPRS and GPS sensors.
The selection of the most suitable processor head after the evaluation of several roller and stroke-based machines, lead to the choice of the ARBRO 1000S. This machine features a stroke movement where the processor body holds firmly the tree, kept still, and the debranching arm moves forward. The absence of rollers enabled the processors to perform a first piling of timber to organize the stacks of different commercial classes and a lower productivity compared to roll processors was not influent, since the extraction of trees by cable crane was indeed relatively slow. In addition, the ARBRO processor performed the same activity (processing trees of the same diameter) requiring smaller prime movers compared to roll processors with lower fixed and variable costs as well as operative cost for the performed tests. Its relatively simple structure and electronics was particularly suitable for the envisioned modifications and integrations.
Regarding the cable carriage, the Tecno system manufactured by one of the consortium partners was selected for modifications while other forestry machines like the automatic chockers and the rope launcher easer have been designed and built from scratch.
As a follow-up of these requirements, the consortium defined the human machine interfaces for the SLOPE platform. The work has been devoted, on a first instance, at the analysis of the current state of the art, followed by the identification of involved actors and their functions. After that, a detailed analysis of the current end user interface has been reported as well as the definition of the final user interfaces for each device through a series of mock-ups for web, mobile and in-vehicle use cases. Adopting the principle of “least astonishment”, the current interaction metaphors have been revamped taking into account project requirements and consolidated workflows to introduce innovation without disrupting consolidated forestry actors’ behaviours.
Attention has been also devoted to the identification of all the available and derived datasets like trees properties, contextual information, harvesting data and wood quality, which could be useful inside the SLOPE forest data model for the achievement of an integrated and interconnected system. All these data have been listed and taken in account for the design of the central database.
The aforementioned actions paved the way for the design of the system architecture in terms of components, technologies and actual deployment. The design made large use of Unified Modelling Language (UML) Schemas to define the interaction between already available components and the missing ones, created following a federated approach. This architecture was not considered static but an evolving structure based on the challenges and the changing requirements encountered during the course of the project.
The requirement study was followed by the development of methodologies and tools to fully describe terrain and stand characteristics in order to plan, optimize and monitor, on-the-field operations. Starting from already existing hardware for low scale remote sensing and multispectral analysis, to reach detailed information on individual trees from high-resolution on-field surveys (UAVs and TLS), the consortium reconstructed and visualized the forest model in an interactive way.
After a first use of Rapid Eye Earth Observation satellite imageries for multispectral analysis where large areas of the terrain have been analysed under different indexes computations (NDVI, CCCI, NDRE, CHL) to detect vegetation type, density and health, aerial (UAV) and terrestrial (TLS) surveys were executed over two specific areas in Trentino, Italy, and on a second one in Annaberg am Lammertal, Austria. UAV flights above the forest using a fixed wing drone equipped with RGB and IR cameras were used to create a 3D dense point cloud (Dense Image Matching Algorithm) of the forest Digital Surface Model (DSM) and, in addition to other information such as Digital Terrain Model (DTM) or Digital Elevation Model (DEM), it has been used to derive a Digital Canopy Model (DCM). This survey was followed by a laser scanning of selected areas (plots) within the test sites, to detect tree properties such as branchiness and diameter and correlate them with the height of each tree. The combination of these statistics with position and height of each tree within a forest stand detected through RGB imagery segmentation and DSM, allowed the build of a virtual forest model where each single tree has its own precise properties and estimated cutting instructions. The work, considering a series of technical and operational limitations, has provided encouraging results.
The constructed model has been optimized for the visualization inside an interactive web 3D modelling for harvesting planning prototype. User requirements and envisioned human machine interfaces have been used to build an online three-dimensional virtual globe based on WebGL technology with high resolution survey data and advanced functionalities for data analysis, operational planning and real-time activity monitoring, such as: imagery and vector data visualization, DSM and DTM data streaming, point cloud visualization of each single plot, 3D forest visualization of trees and logs, measurements tools, cableway and rope launch simulation, spatial queries, virtual marking, cost estimation based on short-time optimization models, report generation and point of interest visualization. From a scientific point of view, high performance rendering of huge number of trees in real time and point cloud support directly inside common web browsers represents an important achievement in the exploitation of state of the art technologies at the service of forestry industry.
Another study carried on for the piloting activities has been the development of a logistic optimization model for a regional wood supply network based on an objective function for the minimization of transportation and other costs. This model supports the decision for the optimal supply network by selecting harvesting plots (origins), intermediate storages (terminals), forest roads and plants (destinations) for different type of products during a period in a specific region. To achieve this goal, analysis of all the relevant logistics locations within the forest for every pilot area, including skyline deployment sites, storage and processing areas, available roads and processing sites in the neighbourhood were performed. Part of this study has been integrated inside the harvesting planning tool.
Based on the acquired knowledge about actors, workflows and hardware requirements, a crucial result obtained during the course of the project has been the design, build and testing of machine prototypes to perform intelligent and smart operations in steep terrains. The following machines have been developed and tested on the field during the two pilot in Italy and Austria, subject to repairs and maintenance actions to improve their performances and cope with unexpected issues: a rope launcher easer, a set of automatic chokers, an intelligent cable carriage, the iTruck system and an enhanced processor head.
The chosen processor head was purchased and subject to a reverse engineering. A study of the mechanical 3D model and hydraulic schemes was then performed with the major effort in the modifications design of cutting forces evaluation, stress wave propagation, scan bar for cross-cut analysis and RFID tags positioning systems. Modifications involved both the processor head and the prime mover. In the first case, for the cutting forces evaluation, the built-in chainsaw was enhanced with linear encoders, flow meters and hydraulic pressure transmitters to detect transversal cutting energy and with load cells and hydraulic pressure transmitters to detect the blades debranching forces. The Near-InfraRed and hyperspectral measurements required the installation of a scan bar, consisting in a mechanical system housing sensors related to NIR, hyperspectral and stress wave (free vibrations) measurement systems. Each sensor was assembled on a movable bar parallel to the chainsaw, placed in a separated area for dirty, lubricating oil and chipping protection. Steppers, hydraulic cylinder, electro valve, inductive switches, protections, mechanical frames and components were used for the handling of NIR cameras, hyperspectral cameras and laser displacement sensors. The stress wave measurement as both time of flight across the log and free vibration measurement was possible thanks to dynamic load cells, hydraulic cylinders, electro valves and inductive switches at the service of mono and tri-axial accelerometers. Both sensors were mounted on two different movable parts, actuated by hydraulic cylinders. The cylinders push the sensors down on the log for the time of flight measurements and maintain the sensors in a safe position during all the other phases. The RFID marking system powering the processor head was built from scratch and mounted in a specially constructed carter in front of the machine. It was composed by a box kept in a safe position during machine operations and placed in front of the log (by means of a hydraulic cylinder) during marking activity. Inside the RFID marking box there was a hydraulic stapling system and an actuated RFID tag roll. Outside the box was placed the antenna devoted to reading each tag placed on the log. In the second case, the Liebherr R310B excavator has been equipped with an industrial PC, a touchscreen, an access point for wireless communications and a CompactRIO programmable controller, installed in the cabin. The work on the processor head has required a complete rewriting of all its basic controls extended with the new SLOPE functionalities. A specific human machine interface has been built for this purpose.
The design and build of the intelligent cable crane carriage on the basis of the technological solutions shaped by the consortium consisted in the installation of an external RFID antenna and a single board microcomputer with Wi-Fi connected with the machine communication system. Studies on the most suitable placement of the Wi-Fi antenna ended with its installation on the lower part of the cable carriage with a hydraulic mechanism to expose the antenna when reading and bringing it in a more secure position when not in use. Specific Data transfer schemes and protocols were defined for CAN Bus communication between the PLC controller and the single board computer and for the selection of in-process parameters to be transferred to the industrial PC installed on the excavator.
New types of automatic chokers interacting directly with the carriage for fast and unmanned detachment of logs after transport have been built and tested. They were equipped with an electromechanical element for moving the unblocking mechanism, an electronic receiver to decode the opening signal and a rechargeable battery. When lowering the load, the cable carriage checks the dynamometer and when the zero value is reached, it directly sends the open command to the chokers which automatically unhooks the load without the presence of an operator.
The rope launcher easer, completely built from scratch from one of the partners, was a system made to launch a synthetic rope from the beginning to the end of a cable line to facilitate the lay out of the main cable. It consists of a wheeled frame housing a power generator and a high-pressure air compressor. A slewing bearing in the chassis base supports the launch tube, which in turn is hinged to the tilt adjustment. Thanks to a GPS system, providing to the embedded computer the end coordinates of the line, the software calculates the launch parameters. The device is also equipped with an electro winch to collect the launched synthetic rope.
Studies on different RFID devices/tags models were performed on operational scenarios within the forest under different environmental conditions, to identify the most suitable ones for the project. Some trees samples have then been used for testing the dielectric properties of moist wood and its capacity to reduce RFID readability. Logs were gradually dried in oven, repeating RFID reading under the same conditions with different moisture content of logs, from maximum water content to oven dry. Since experiments confirmed their reliability, a new RFID UHF tag design was commissioned to the most suitable provider to be used for manual and automatic processor head tagging.
To bring intelligence to the transport vehicles for timber and biomass within the project scenario, providing real-time position, weight and fuel consumption for route and costs optimization, a hardware and software prototype called iTruck, was built. The prototype, made up of a single board computer, with GPS, Wi-Fi, 3G and Bluetooth connectivity, ensured recording and transmission of data to the central information system as well as logs traceability as it was able to connect with portable RFID readers used to scan the loads before transport.
A data storage back-up and management solution has been studied and built to handle data coming from on-the-field sensors, consisting in procedures and software running on the excavator industrial computer, to acquire, process and transmit data to the SLOPE central database.
In parallel with the development of the intelligent machines, a methodological basis for a semi-automated and real-time grading system for forest production was studied and applied on the processor head. Achieving this goal required an extensive testing and tuning of different sensors within the laboratory and on the field to ensure an optimal definition of thresholds and measurements for the final log grading. Grading is an important step in the production chain, when the resource properties and the quality of the product are assigned. Proper grading process leads to accurate definition of the timber value on the market, and in some cases, an extra value can be obtained for the timber products. Grading logs in forest is the first assessment of the technical quality of wood and it is a starting point for following actions in the fields of timber trade and its valorisation. Near-infrared and hyperspectral sensors have required more effort as the design of their mounting on the scan bar has been particularly complex. A proof of concept of a scan bar has been developed but its functioning was not tested on the field due to its fragility and the very harsh forest condition especially, on the Austrian forest site. However, promising results have been demonstrated in the lab using the scan bar on log slices taken directly from the field.
The SLOPE consortium after an evaluation of the state-of-the-art of the current practices and rules used for logs grading (user’s practices, legal norms, standards, regulations), defined five quality grading strategies to be tested, almost entirely developed using LabView software.
A 3D quality index, making use of laser scanners for 3D mapping of standing trees, was used to determine tree shape and stem dimensions (including diameters, taper, tree height, etc.). A custom software capitalizing on the partners’ experience was defined for this purpose and to simulate optimal cross-cut assuring the highest gain of the forest owner and compatibility with the local market demands. Results showed that basic log indicators could be easily measured using the stem 3D model created using TLS data with an estimate of the current indexes for each stand: waste ratio, products ratio, total profitability index and harvesting loss risk. Instructions and tree logs models saved on the central database were always accessible through the 3D planning and monitoring system.
A Near-InfraRed spectroscopy quality index was used for the determination of selected wood properties and defects. Chemo-metric algorithms for the prediction of intrinsic quality indicators were developed and validated using different types of sensors. Studies on the effect of wood moisture variation on the NIR spectra were performed on laboratory equipment and experimental data was used for extensive calibration. Experimental results highlighted the MicroNIR Spectrometer as the most accurate and suitable for the installation on the scan bar. NIR spectroscopy demonstrated to have a great potential for assisting quality grading of logs although there was no evidence of its application in the forest on the literature before SLOPE.
A hyperspectral imaging quality index was used for the characterization of bio-resources along the harvesting chain. As an extension of NIR over a whole area instead of a single line, this type of index was calculated through the development of chemometric grading models for qualitative analysis of spectra and a number of laboratory trials on wood disks taken from the field activities during the second pilot. Accordingly with the literature, hyperspectral imaging has a great potential for quality grading but its application within the forest was never been experimented. Trials lead to the design of a rig of 16 Hamamatsu sensors for the deployment on the scan bar.
Experiments validated these two approaches, as sensors were detecting the unique material properties of logs in objective and repetitive way, but the number of tested logs was not sufficient to assure precise quantitative analysis (such as direct computation of the chemical components content or detection of defects not previously included in the chemo-metric model).
The optimization of log/biomass analysis through stress-wave acoustic measurement was another evaluated measurement protocol. It was based on the observation of wave propagation and its correlation with mechanical and physical properties of the wood material (i.e. dynamic modulus of elasticity, density) as well as the presence of damage, decay or other specific defects. Although this approach was not new in the literature, the combination of multiple stress-wave measurements in a unique testing setup, the evaluation of different grading-dependent factors (e.g. spiral grain, knottiness) and integration with data from other sensors (e.g. estimated density values) constitutes a new research approach. Results showed that time of flight as well as free vibrations may be highly useful for quantification of log quality. However, the range of log characteristics, was still not sufficient to assure a precise quantitative analysis (such as direct computation of the value for modulus of elasticity or mechanical strength), but the quality indexes might still be applied for the screening of logs before shipping to the final user.
A cutting forces quality index has been defined with the analysis of two types of forces: cutting resistance of the chain saw cross-cutting the log, measured through hydraulic pressure and oil flow which are affected by the wood density and cross section geometry and resistance of the debarking knives during the stroke, measured by hydraulic pressure and mechanical deformations of the knife holder, which are affected by tree knottiness. These techniques, already used in secondary wood processing, represent a pioneering approach on the field. Results showed that measuring cutting power while cross cutting logs as well as when debranching, is highly useful for quantification of the quality and can provide additional resources not available in other systems.
All the aforementioned sensing techniques were used from the SLOPE consortium to develop a prediction model fusing each single index into a final quality value for the log/biomass that is being processed. Three approaches were tested: An expert system based on the “if-then” rules, multivariate analyses and statistical models (i.e. principal components analysis, partial least square regression, soft independent modelling class analogy and cluster analysis) and neural networks with learning supervised by an expert grader. After the tests, a linear weighted combination was finally selected as a data fusion algorithm, allowing customization of the properties importance for any quality sorting scenario through the setup of values for each quality index.
The development of forest machines and automatic quality grading strategies proceeded in parallel with the definition and development of the mountain forest information system (FIS), storing the data related to the harvesting process and interconnecting all the services available among the partners, creating a unique entry point for all the wood processing operations to the several involved actors (planners, logistic operators, brokers, end-users, forest owners). A comprehensive database was designed for these purposes, supported by remote web services to support the other SLOPE modules and clients for real-time control of the supply chain, short and long term planning as well as monitoring and selling of wood.
A number of software and hardware components working on top of the SLOPE FIS web services were developed for near real-time control of operations to provide planning information, storage optimizations, reduction of delays and advanced reporting taking into account critical conditions in the forest (network availability, rough environment, etc.). A first mobile app able to connect with handheld RFID reader and external GPS, was developed for field operations and database review, supporting tree RFID marking, tree/log status change for traceability, manual quality grading and forest stand analysis. The cable carriage internal software was modified to record tree RFID tags during the loading phase and record statistical data about the machine during its activity and transmit it to the black box for data backup on the excavator. The excavator and processor user interface and control software were re-engineered and rewritten in LabView to support prototype commands modifications and expose to the operator specific harvesting parameters. When the tree is collected with the processor head, the system shows its identifier, RFID tag, timestamp and suggested cutting instructions to produce logs while during the production of each log, the system shows length, diameter, volume, weight, branchiness and its quality index. A second mobile application has been also defined to create and manage storage areas and movements of logs assortments between them, while the 3D harvesting and planning system was updated to support monitoring of the status of a harvesting operation (storage areas, felled/processed trees statistics etc.).
The ecosystem for real-time control of operations has been the perfect case for additional studies of short-term optimization planning techniques. This process required an analysis of the best algorithms for cost and productivity calculations of several tasks involved in the wood processing chain, like transportation, felling, cable line setup, etc. minimizing storages, evaluating biodiversity and forest integrity impacts through a set of indicators (average slope, tree distances, height from terrain, surface area, obstacles, etc.). A selection of these models (cable way setup time, tree felling time, route transport costs) was implemented inside the 3D harvesting planning software to support forest planners in the estimation of times and costs of on the field operations. Values were computed in real-time based on selected terrain, surface and tree properties and could be exported as a report for on-the-field validation.
Medium to long-term optimization models were added to the SLOPE FIS for strategic and tactical planning of forest stands on a longer timeframe, providing instruments to optimize the use of available resources to satisfy the market demand of wood in the upcoming 5-10 years. This has been possible through the deployment of a software connector between the SLOPE FIS database and the IPTIM long term planning service. Forest owners can send their surveyed SLOPE stand data into the IPTIM service and receive a forest stand long-term management plan (i.e. felling, mechanical cleaning, soil preparation, pine planting, mechanical grass vegetation prevention, etc.).
The consortium integrated these long-term functionalities in a system for online purchasing and selling of wood products (timber and biomass), allowing identification and tracking of materials, negotiation, bidding, analysis and pre-selling procedures. The online portal called Wuudis is now public and available on the market as a full product, extensible with the SLOPE services once ready for the market.
To redefine the forest production workflow, surveys data, intelligent machines, software and hardware for real-time control and resource planning required the achievement of a complete integration of the different elements composing the project platform. For this purpose, the project envisioned an incremental integration plan having the FIS services and the forest common database model acting as the central communication hub between each platform component, with a web portal as the entry point for all the actors involved in the wood processing chain. A hybrid integration and validation approach, based on waterfall and agile-like development was executed. Three steps of integration with growing complexity have been done, each one with a sprint every two months and test cases validation followed by further developments.
The first integration phase, between the forest inventory and the harvesting systems focused on the access to the forest data model from all the hardware instruments, tracking systems and planning/visualization tools. Web services and the components accessing them were tested in depth to reach a percentage of satisfied use cases of 70%. The second integration phase, involved the forest management software. On-the-field data was connected with the central forest information database and used by the short/medium/long-term data analysis and synthesis services. Testing was executed to reach a threshold of 80% of completed use cases. Finally, the third phase devoted its effort for polishing and refinements to reach a test case completion percentage of 90%. Overall, a total of nine rounds of testing and validation were performed during the whole project to ensure the proper integration of the forest inventory with the harvesting system and the forest planning module. Satisfied test cases reached a percentage of 88%, a value not far from the target and affected by a missing set of functionalities foreseen at the beginning of the project for the ERP module but not implemented as already available inside other components. In parallel with the integration, each partner worked to refine and improve its hardware and software to reach a unified workflow. As an example, the 3D harvesting and planning tool added and improved the following functionalities: workspace save and loading, automatic generation of multi-language legend from maps, report creation and download, visualization of detailed stem data, tree virtual marking, spatial query and statistics visualization and point of interest visualization. All the software components on machines, mobile devices and fixed computers were able to communicate with the forest information system.
Three progressive techno-economic analyses of the SLOPE solution elaborated during integration and validation on the field highlighted its competitiveness with respect to the current techniques. Although a number of necessary assumptions, cost estimates shown a negligible difference of 1% between the two workflows approaches. The additional SLOPE surveys and processing costs were balanced with the removal of logs sorting costs, now directly performed during harvesting activities.
Two on the field pilot tests supported the aforementioned integration and validation activities with a major focus on the hardware of the intelligent machines. In a first instance methodologies were defined to evaluate the achievement of SLOPE during the pilot experiment, based on a detailed analysis of current systems by drawing process flow charts, identifying bottlenecks as well as possible strengths and risks of the supply chains. After that, responsible partners worked on the preparation of pilot areas.
The SLOPE project detected two valid tests sites in different period of the year and environmental conditions: north of Italy in summer and Austria in autumn. The activities started earlier than planned during the first year, with the UAV and Terrestrial Laser Scanning surveys on the Italian test site, while the other machines were tested only during the last year of the project. For the first demonstrator in Sover, Italy, the consortium marked trees with traditional and experimental techniques (RFID and tablet apps), identified the cable yarder landing area, the potential storage areas and concluded agreements with the forest owner and the forest/harvesting company for the operative aspects of the demo. Operational procedures, legal requirements (CE limits), safety instructions, felling the trees, installation of the cable yarder and carriage were studied before the actual deployment, updating time consumption, effort, productivity and costs. A similar process was carried out in Annaberg, Austria with a major complexity due to the logistics of bringing machines (whole system Tecno carriage, the excavator with the processor head, radio controlled chokers) located in Italy to the pilot.
The first pilot demonstration was performed in July 2016 for a total of one week of tests in parallel with regular harvesting activities. The SLOPE excavator processed a small number of trees while a roller processor head handled the majority of the assortments. The enhanced cable carriage worked continuously for the entire week with only one day of stop due to an issue in the power supply system for the single board PC. Collected and transmitted data were in line with the expectation. The processor head produced less data results than expected due to several issues encountered on the field like the overheating of the excavator or the required manual triggering of sensors, but was crucial to plan the improvements for the second demonstration.
Tests in Austria, in October 2016, ran more smoothly, even under bad weather conditions (snow), thanks to the automatization of sensor analyses and the revision of the excavator control system. All the planned 18 trees were extracted and processed in just two days and from them, eight logs were completely scanned and graded for quality. Overall, the processing speed was drastically improved and valuable experimental data for log quality grading was produced for further analyses. Preliminary grading results, in comparison with expert evaluations showed that the SLOPE system and the visual system are generally in good agreement, but due to the low number of processed logs the statistical confidence of the results was limited.
Compared to the visual subjective grading system, the SLOPE technology aims at being well-defined, objective and repetitive, trying to avoid subjective bias. However, reliable correlations between the sensor data, mechanical wood properties and its potential uses still needs to be developed to reach this goal. Well-calibrated sensors and models and the scientifically based definition of threshold values for the different parameters and their combinations are a pre-requisite for the quality rating. Once the system will be calibrated and the models established the openness of the SLOPE grading system will enable the definition of further categories of resource uses along the production chain by improved description of material properties.
On the field trials were fundamental for the improvement of intelligent machines, making them, in many cases, more than just a proof of concept. Full wood traceability was achieved and preliminary tests on automatic grading showed promising results besides more tests on the field would have been advisable.
The pilot activities were supported by a specific set of training material to be used on and off the field to transfer knowledge, skills and competences built during the project, to stakeholders and on-the-field workers. The consortium published a series of 15 video tutorials for the usage of the SLOPE online portal and performed demonstrations to forestry experts. The consortium planned also a set of communication tests for the cable carriage before the pilot deployment and published multi-language customer satisfaction questionnaires, several training meetings for the operators to explain the different work systems and prototypes, prior to the pilots. Additionally, the consortium organized and delivered a training course on the use of remote controls and load/unloading instructions with the Tecno cable carriage to the company in charge of the harvesting on the first pilot.
SLOPE performed a significant number of activities ensuring the dissemination of the described project goals to a wider community. This included publication of articles, organization of workshops, and presentations in conferences. Currently, the most important showcase for the project is its website (https://web.archive.org/web/20190818085734/http://www.slopeproject.eu/) constantly updated with all the relevant information, including the related events. A number of parallel web 2.0 communication channels have also been deployed and constantly updated with the aim of ensuring potential stakeholders’ involvement in the project. The consortium has been also active through a number of dissemination activities including scientific publications, meetings and organization of SLOPE-branded events as well as presentations of the project to national or international conferences.
The SLOPE project results and the acquired forward knowledge have been discussed and carefully analysed from the consortium after the second demonstration producing an exploitation and business plan and a shared licensing strategy.
Potential Impact:
SLOPE developed an innovative and cost-effective technology platform, providing a solution for planning, managing, measure and monitor wood extraction and production in mountainous areas. This is expected to have significant impacts in various areas.
From a scientific point of view, an integrated platform, a forest data model, novel technologies for forest data collection and reconstruction, 3D web visualization and planning tools and mechatronic improvements of harvesting machines will open new routes for the optimization of the wood supply chain.
From an economic point of view, the project will drive new products and services from partners to potential stakeholders interested in the exploitation of their foreground knowledge. Wuudis start-up and its marketplace or the rope launcher easer, soon on the market, are just some examples.
The current market demands raise a number of exploitable opportunities for the SLOPE technologies. Requests for sustainable forest planning and management is increasing within the global forest industry and forest managers need to find the balance between society's increasing forest products demands and societal benefits, and the preservation of forest health and diversity. Sustainable management of forest resources includes determining, in a tangible way, how to use these resources today to ensure benefits, health and productivity in the future. In this context, harvesting planning and optimization is an important foundation for sustainable forest management. This results in a growing demand for forest resources measurement and monitoring during the life cycle of the forest. However, the use of traditional methods has been proven not always cost-efficient for these new challenges. The main constraints for forest managers and landowners is the cost of the labour-intensive collection and management of the data and the activities related with the harvesting and extraction of the timber. The use of new techniques should address these issues.
A first opportunity for the SLOPE project is a better operational harvesting planning which is required in most European countries before proceeding with the harvesting activity. Operational plans are developed for each individual harvest area, based mainly on on-site inspections. Maps of the harvesting area, showing a detailed activity plan are required. Major points to be covered in an operational plan include the identification of:
• Harvesting area (usually 50-100 ha), location and boundaries (following topographic or natural features)
• Areas to be excluded from harvesting through prescriptions for flora and fauna protection, water quality protection or other identified reasons
• Silvicultural prescriptions to be adopted for different forest types
• Methods of tree marking for selection and protection
• Volume of wood to be removed by species and size classes
• Location, design, construction, maintenance and closure of roads, landings, log ponds and skid tracks to minimise disturbance to forest, soil and water resources
Current methods are based on a significant labour intensive work including field inventory, use of forest model analysis, technical GIS skills to create maps and harvesting mangers for the final decision making. Additionally, when using cable crane in mountain areas, engineering skills are required to plan the correct set up of the crane and other machineries in the field. Outputs of this work are usually documents with a set of maps including the location and design of the different infrastructure and silvicultural prescriptions. These outputs need to be shared within the different people involved in the project via email or paper maps.
Another opportunity consists in the provision of new tools to facilitate the different analysis by automating the forest inventory analysis, GIS analysis and planning of the cable crane as well as to combine these tools in a single platform (currently several software are used to process this information), enabling virtual collaboration of the different staff involved in the collection and analysis of the information required for the harvesting plan.
On top of integration, automation and sharing of harvesting plans, there is also an opportunity to improve the quality of the inventory data used as a base for the decision making. Currently it is mainly based on measurement of trees providing a rough estimation of their volume, based on models. The use of new techniques such as Terrestrial Laser Scanner (TLS), allows the creation of accurate 3D models of each single tree, providing a better estimation of the forest resources and facilitating the decision making.
These opportunities are supported by a clear trend of increasing harvesting operations. New regulations necessitate a more and better amount of information regarding harvesting plans in most countries around the world. As the technology evolves and social concerns about deforestation increase, harvesting plans need to be more detailed. The volume of timber harvested annually in Europe is 735 million m³ and 756 million m³ in North America, with a worldwide total of 3,359 million m³. Most of these harvesting operations are planed carefully from the start, representing a potential opportunity for better harvesting planning systems. A study based on FAO’s FRA 1990 dataset (updated to 1995 in 1997) estimated that there would be a considerable increase in the global round-wood supply from 1.800 billion m³ in 2000 to 2.275 billion m³ in 2040, which would largely be met by production from plantations. Although the productive forest plantation area was estimated at only 3 percent of the global forest area, in the year 2000, the study predicted that plantations would meet 35 percent of the global round-wood supply in that year, rising to 44% by 2020 and 46% by 2040. These figures shown a clear growing trend linked to even greater increase of the requirement for more detailed harvesting operation plans using new technologies.
A fourth opportunity resides in the timber tracking. New regulations (e.g. 2010 EU Timber Regulation) and certification schemas (FSC, PEFC) require that forest owners and managers ensure the wood was coming from where its suppliers claimed. Reducing illegal logging could substantially increase revenue from the legal trade in timber and halt the associated environmental degradation but law enforcement and timber traders themselves are hampered by the lack of available tools to verify timber legality. The use of new technologies such as RFID systems integrated in forest machines, transportation fleet and forest workers (chainsaw operator, forester), can offer a unique capacity to track and trace both trees and timber logs during the different stages of the supply chain. This technology will help forest owners and managers to meet their requirements of traceability regarding legal requirements and timber certification. At present, the main certification schemes, PEFC and FSC, cover a forested area of 236 Mha and 183 Mha respectively worldwide (in Europe the certified area is 95.5M ha and 81.8 Mha respectively). Therefore, there is a potential opportunity that a significant part of this certified forest will adopt automated traceably systems.
A fifth opportunity is related with the sales of timber. Wuudis Service, developed in parallel with SLOPE project, is a mobile and web service platform for real time forest management, care work and harvesting monitoring as well as online purchase and sales of timber, enabling detailed documentation and tracking of purchased wood, going from forest stand down to a compartment level. Wuudis is also equipped with features and functionalities for easy data sharing between all stakeholders in the value chain and for certification schemes.
The final opportunity is represented by the adoption of innovative machineries. In forestry, despite the increasing automation, is difficult to create a supply mechanization chain because forestry areas are always different and changing. SLOPE created a communication system among machines, an automated supply chain where man can rely on artificial intelligence and has a control task to perform plants evaluation, morphological definition, electronic marking of trees until harvest, weighting, calibration, the creation of a classified assortment and logistic transport. These machines have the potential to be competitive product for the market and some of them are planned to be included in the partners’ product portfolio of 2017. More specifically, the enhanced cable carriage will be equipped with a series of accessories such as the electronic scale, the tilt sensors, the optimization system of consumption and consequently pollution abatement, the back-up data system with Wi-Fi for the collection and storage of daily production data. The tag reading system with retractable antenna and automated radio control for the release of chockers will be offered as optional. The expected customers’ feedback would be the immediate idea of a carriage that, as well as having an excellent efficiency on the automatic work, provides an overview on the status of the forestry installation, on production rates and status of works. Automatic chockers are going to be launched on the market by mid-2017 after further development to make them lighter and easier to use while the rope launcher, the most welcomed product across the project is currently under certification (CE Mark) and is expected to be built at the end of the year or at the beginning of 2018. Its potential low selling price elicited enthusiasm and interest and many users’ operators have already contacted sales office with purchase requests.
From a dissemination point of view the work performed by the entire consortium has been devoted to the submission of scientific research articles participating to international conferences, project follow-ups proposals, submission of new dissemination contents to the official website and social media channels (YouTube, Facebook, Twitter, LinkedIn, SlideShare) and featuring on other online websites as well as keeping contacts with potential stakeholders.
Overall, 28 scientific papers have been published on national and international journals and conferences, with the consortium attending more than 30 conferences and more than 25 trade fairs. Four techno-scientific workshops and one final conference followed by an additional final workshop were organized. Results were disseminated with four press releases, five newsletters, two flyers, one infographic and more than 200 updates on social-media channels with 23 videos and 134 online slides presentations.
The project planned the involvement of industrial partners inside the consortium with the constitution of an advisory board (IAB) consisting of 3 senior experts, nominated and invited to assist to the demonstrations on the field. Two members out of three were in position to attend the first pilot and technical meeting in Italy, in July, while the third member managed to attend the second pilot in Austria. The board, after each meeting, provided valuable feedbacks on how to develop and further exploit the integrated SLOPE system results. An initial work was carried out to contact standardization technical committees (CEN/TC) and other public authorities to introduce them the first preliminary SLOPE results for quality grading.
In addition to the noteworthy amount of resources spent for the communication of results to potential stakeholders, the consortium worked on the definition of a business plan before the end of the project, using the exploitation booster service brought by the European Commission to FP7 and Horizon 2020 projects. For this purpose, a virtual meeting with business expert consultants was organized in December to define projects key exploitable results (KERs) and design their lean canvas models for future project evolutions. The outcome of this meeting has been a list of seven products and services with an accurate evaluation of innovative aspects and unique selling points, briefly introduced below.
Product/Service 1: Detailed digital model of forest viewable by many parties
Through the first service, large areas of forest can be surveyed in a small amount of time which is constantly reducing due to technology evolution. Aerial data capture is easier and more cost effective making it suitable for adoption and sharing with planners, foresters, and prospective buyers. With the innovative systems introduced by the SLOPE project forest owners can easily acquire more detailed data about their forest property in the inventory/planning phase. This service makes possible to generate a very detailed dataset with a cost-efficient system. UAV images are used for a first screening and characterization of the forest areas and, combined with in-field laser scans can generate a 3D forest model. The obtained model includes a highly-detailed canopy and terrain definition, as well as a full inventory of standing trees and the potential timber assortments in each single stem (bucking prevision). This result constitutes a feature unique in the market. Expected time to market is planned within six months.
Product/Service 2: Forest Planning Platform with 3D forest models
The second product is the real-time interactive 3D geographical visualization system. Using forest models acquired from UAV/TLS it helps in the planning, simulation and monitoring of forest production activities. Features include: 3D forest model visualization, open data, cable crane setup, slope analysis, measurements, working area setup, truck routing as well as spatial queries on the forest for timber product breakdown, cost analysis, cost forecasts and reporting. This service includes a completely web solution, globally scalable, supporting different forest actors with a level of interactivity unique on the current market. This can be used for a "virtual marking", identifying the trees to be felled according to their position, characteristics and commercial value. The model can be integrated with all the available digital data, such as property limits, road network (public, forest, etc.), protected areas, soil types, etc. Such detailed information, provided in an easy-to-use interface will return a reliable estimate of the stumpage value and harvest costs of the forest. It will also allow the owner to better understand the real value of its property, as well as facilitate the development of new forest management plans or update the existing ones. The model can also assist the chain of harvesting operations helping the definition of stumpage value either for selling the standing forest or for contracting the harvesting service, reflecting almost in real-time on the field activities (e.g. forester inspections). Expected time to market is planned within six months.
Product/Service 3: Automatic tracking and traceability system for timber products
The third solution consists in the integrated system for manual and automatic management and reading of RFID tags. In a market-adaptability focus, this system may provide the following services singularly or in a unique service:
• Tracking of single tree/log items and bulk goods along the supply chain
• Trace timber products with full details of location and date of key actions (felling, extraction, transformation, transportation)
• Transfer quality parameters and instructions for optimal handling/transformation by means of the same RFID system, maximizing added value (e.g. optimal value recovery)
This solution is the first auto-ID system integrated in forest machines, transportation fleets and forest workers combining widely available RFID technology with the devices and machines used during the raw timber supply chain, creating a unique system, accessible online and capable to track/trace both trees and timber logs. Most of the system is ready for the market but six month of further development and a partnership with a processor head manufacturer are needed to improve the solution robustness.
Product/Service 4: Wuudis mobile and web service
The fourth solution is an independent and neutral mobile and web service called Wuudis marketplace, including forest management, timber and biomass online sales and purchase service, and also a bidding service for forestry care works. This platform will cover also enterprise resource planning (ERP) for contractors in the next year. Currently, competitor’s solutions are closed services targeted only to their own customers/members. Wuudis is provided as free service for forest owners, requiring a commission fee to contractors or timber and biomass buyers only after a successful deal. Wuudis solution is already available on the market.
Product/Service 5: Sensorised processor head with wood classification grading
The sensorized processor head with wood classification grading is the fifth key exploitable result. An intelligent machine to provide automatic wood classification grading during tree processing, a unique feature in the market. The grading capability reduces uncertainties and errors in manual wood quality grading, providing real-time evaluation of wood under several different indices. Knowing in advance the wood quality, allows the owner to maximise its profits. Further developments and a partnership with a processor head manufacturer are needed to finalize the prototype.
Product/Service 6: Rope launcher for timber extraction
The sixth KER is consisting in the system to launch a synthetic rope from the beginning to the end of a cable line, to facilitate the placement of its main cable. The system helps reducing the number of hours required to setup a cable line as well as the physical effort for the forest operators increasing the productivity and the margins. Despite the relative technological simplicity of this machine, its concept is unique in the market. This is a brand-new product that can save days of work when compared with the current process of manually bring the heavy rope uphill. The product is ready for the market but is currently missing the proper certifications and is expected to be sold by the end of 2017.
Product/Service 7: Automatic chokers to unhook timber logs
The final exploitable result consisted in the automatic chokers to allow the operators to unhook the logs without the need of getting close to them, staying in a safe position. These automatic chokers present the advantage of directly interacting with the control board of the motorized cable carriage, directly sending the open command to the chokers only when there are no forces detected by the dynamometer on the cable, increasing productivity and safety. Expected time to market is planned within six months.
To understand how the vertical and horizontal products and services could work individually and, the consortium worked on a set of preliminary lean canvas models, to identify joint-exploitation plans, involved partners, key metrics, unique value proposition, communication channels, customers, early adopters, costs and revenue streams. Although the main focus was on individual exploitation of results, a SLOPE global platform as a services marketplace has been envisioned, acting as the potential integration space of the whole value chain, offering a unique shared economy model, linking in the future services such as Wuudis with other ones such as the 3D forest model, tracking system and intelligent processor head. Ownership was assigned to the different partners that may run one or several services based on the SLOPE platform while producers/suppliers were identified in forest managers, harvesting companies and forest public bodies. Finally, consumers creating the necessary demand, were identified as forest owners, harvesting companies, timber buyers and certification authorities.
Since the type of industrial research activities carried on by SLOPE could have an impact on future forest production, all the elements used and generated by the SLOPE project had to be protected by the appropriate IPR and licensing policies. Therefore, the consortium worked on an IPR document to identify each element used or generated by the project that can be under a license and to define which kind of license should be applied. Both hardware, software, data and documentation were considered and reported.
The SLOPE project did not constitute only a joint effort to improve the current forest production workflow, but also a mean for each involved partner to increase its competences and expertise to be more competitive on the market. This process went beyond the foreground knowledge acquired in forestry and was transferred within each partner’s individual field of expertise.
Partner GraphiTech, was able to develop new multi-purpose interactive GIS spatial functionalities from the 3D harvesting and planning tool, such as: measurements, planning, multi-resolution terrains, routing, OGC and Open Data support, spatial queries and point cloud visualization. The partner plans to transfer them to its commercial GeoBrowser3D (www.geobrowser3d.com) platform as a new set of features for the market for: advanced data management, terrain and custom layer visualization, planning and collaboration, point cloud and BinaryGLTF visualization, massive rendering of objects and OGC compliance. It also plans to further develop and exploit the 3D Harvesting Planning tool, advertising it through the GeoBrowser3D website, social network and contacts through its network targeting public administrations, urban planners, companies that deals with logistic of any kind of goods, Geo-ICT companies and tourism agencies listed inside the SmeSpire (http://www.smespire.eu) network database.
Partner CNR, was able to develop and test in-forest grading systems (processor prototype) for early definition of timber quality as well as develop and test an auto-ID timber traceability system based on RFID tags. As a research centre, it does not plan to go to the market, but to use the aforementioned results to further test the intelligent processor and approach the dedicated industry in search of a partnership and to work on a collaborative project proposal with PEFC in Ireland, Spain and Italy, involving forest owners and managers.
Partner Compolab finalized the development of an intelligent processor head optimized for working at cable crane unloading sites and able to perform a series of analyses on the processed timber, as well as marking each log with RFID tags reporting and storing the collected information. Besides some earthmoving machine operator license, Compolab acquired new engineering design skills for heavy machines, hydraulics, sensor electrical integration and conditioning, traceability systems, extending its network of contacts in the wood industry, harvesting manufacturers, forestry end-users and stakeholders. As an engineering office, Compolab plans to carefully evaluate the submission of patents in the field of automatic log marking systems and cutting forces estimation and to establish cooperation with harvester manufacturers providing their expertise for the integration of their developed sub-system on existing machines and the design of novel machine equipped with marking systems and/or grading capabilities.
Partner Coastway being relatively new to the use of UAV in forestry learned from SLOPE that there is a large market for their service but the technology is changing rapidly and they have to stay up-to-date with recent advancements. Additionally, pricing is strongly influenced by the covered area and flight authorization require time and effort to be obtained. Capitalizing on the experience of SLOPE, Coastway is moving into new markets in Scotland and Europe convincing landowners to use a combination of TLS & Aerial surveys, moving into fields beyond forestry such as environmental monitoring and coastal erosion. On this regard, Coastway has been invited to Horizon 2020 Project “Life”.
Partner MHG, during the project, developed a prototype for online purchase and sales market place which was extended to a commercial service under the Wuudis® trade mark, creating also a spin-off company for this purpose, called Wuudis Oy. Additionally, MHG build a commercially ready interface with Simosol/IPTIM for forest exploitation optimization and is involved in new project proposals with some consortium partners. During the project, MHG learned new IT technologies like AngularJS web framework, Ionic mobile framework, REST API based stateless application architecture and Jelastic: container based cloud environment. Its plans for the future are to pilot the launched Wuudis Service through the DataBio Horizon 2020 project in Finland, Czech, Belgium and Spain, searching sales partners in these countries, offer joint added-value services in partnership with Finnish forest planning service providers, apply for further Horizon 2020 SME programs and closing a first round of investment for Wuudis Oy.
Partner BOKU, during the project had the opportunity to test innovative machines and equipment in real operating conditions evaluating practicability, efficiency and acceptance and developing an approach for prototype evaluation, optimization models for wood logistics, quality control strategies via Near-InfraRed and Hyperspectral Imaging and collaborating with partners from other sectors than forest engineering. As a research centre, it is mainly interested in scientific results collaborating with other partners from research and industry for projects, publications, courses, training schools, workshops etc. For these reason, future plans include development of logistic tools, standards for evaluating machines and prototypes as well as technology transfer in steep terrain timber harvesting, usage of developed tools and models during student’s courses and work on new research projects.
Partner FlyBy was able to develop new software components for Earth observation data download and automatic processing for vegetation classification and continuous analysis through NDVI and chlorophyll indexes. Additionally, the partner deployed web services ready to supply geospatial products to the end users and was able to build contacts with new forest researchers and end-users. Plans for the future include the use of developed components as services for final users in forestry and agriculture (i.e. vineyards monitoring) as well as exploit the contacts achieved in the framework of the SLOPE project to develop new projects and new business.
Partner Greifenberg improved and created new forestry machines during the project that are almost ready for the market. Its plans are focused on bringing them to the market to speed up forest operations and security.
Partner Treemetrics defined through the course of the project a more integrated 3D forest model, with a new tree quality index, new use of the TLS 3D models for harvesting planning and new use of the Forest Warehouse 3D cutting instructions within the SLOPE system. Additional development was carried on for their mobile application including new functionalities like: RFID support, tree status update, better integration with remote systems, connection with SLOPE central database and integration of Treemetrics database with the one in SLOPE. Future plans include the provision of data to the SLOPE services, like the forest inventory data and the 3D model of the forest as well as support for on-the-field marking activities.
Finally, partner ITENE was able to build a new tracking solution using RFID technology, discovering a completely new industry sector without previous experience in the forestry industry. Future plans will be focused on continuing with the development of their traceability solutions within this new industry, making also use of collaborative funds like the PEFC Collaboration Fund together with some SLOPE partners.
List of Websites:
Website: https://web.archive.org/web/20190818085734/http://www.slopeproject.eu/
Showcase Video: https://youtu.be/Ml8nQzf17Mc
Social Channels:
Facebook: https://www.facebook.com/pages/SLOPE-Project/351131505025437
Twitter: https://twitter.com/SLOPEProject
SlideShare: http://www.slideshare.net/SLOPE_Project/presentations
YouTube: https://www.youtube.com/user/SlopeProject
Forest production in mountains has a significant role in the world economy, being these areas 23% of the total Earth forest. Projections suggest that this trend will not change before 2030 with the major production growth in Europe, Russia, Eastern Europe and South America. This growth involves all the major wood sectors like wood-based panels, paper, industrial round wood and bioenergy and is a consequence of increasing energy demands interest on wood properties and in the construction sector. However, wood production in slope conditions differs from traditional one for the limits imposed by terrain morphology, for complex operation planning and hardware requirements, like the use of manual chainsaws and cable cranes instead of processor heads adopted in flatlands. Powerful and intelligent machines must be developed for forest operations in steep terrain together with improved types of workflows.
The SLOPE project fills this gap by developing a new process and an integrated information system for the optimization of forest production where remote sensing (UAV and Satellites Imageries) and on-field surveying systems (Terrestrial Laser Scans) are used to build a detailed 3D forest model to support the entire wood process, from tree RFID tagging to cut and debranched logs sent to the sawmills.
The adoption of this model, enables forest data sharing among the involved actors through mobile devices and complex harvesting machines, improves wood output, adapting the available wood quality with the market demands, reduces manual work, increases productivity and speeds up monitoring and quality assessment thanks to detailed interactive and always accessible planning tools.
Enhanced machines based on market available devices have been built to analyse and track in real-time trees and logs. An intelligent processor head, equipped with state of the art sensors identifies the marked trees suggesting how to optimally cut it into logs and evaluates in real-time their quality during cutting and delimbing phases, combining the index quality result coming from different measurement technologies. Smart RFID tags track trees from the forest to the sawmills, with readers mounted on foresters’ roughed tablets, cable crane systems and smart-trucks, while an enhanced self-motorized carriage equipped with new remotely controlled and programmable automatic chockers increases the operational security.
Long-term planning as well as auction and selling optimization has been achieved through the connection of the digital forest model with state of the art forest planning systems integrated inside a wood trading portal, launched on the online market during the course of the project as a first example of the projects’ exploitable foreground.
Extensive trials have been performed on two test fields under different environmental conditions to demonstrate the effectiveness of the entire system and improve its reliability, opening the door to future results exploitation.
Project Context and Objectives:
Mountains occupy about 35% of the European territory and are mostly covered by forests. Forestry operations in mountain areas are seldom performed by fully mechanised systems typical of the industrial sector. The majority of operations is still characterized by manual felling and extraction of timber by cable cranes. In fact, due to the limits posed by steep terrain, non-uniform vegetation, long vegetative succession, poor road networks, hydro-geological constraints and limited storage areas on the field, harvesting and extraction are more expensive and less flexible compared to the cut-to-length systems based on wheeled machines, commonly used in flatland forests of European Nordic countries.
In order to be competitive, powerful and intelligent machines must be developed for forest work in steep terrain. The SLOPE project integrates information from remote sensing and on-field surveying systems, to support analysis for the characterization of forest resources. Spatial information can be integrated with multi-sensor data in a model for sustainable forest management and for optimization of logistics during forest operations.
The mission of SLOPE is therefore to develop an integrated system that increases the efficiency of harvesting and logistic operations and optimizes timber production in mountain areas reinforcing the communication between its actors and filling the existing gap between more expensive, less flexible forestry operations in mountain areas and optimized cut-to-length systems, commonly found in flatland forests. For this purpose, spatial information and multi-sensor data from remote sensing, UAVs (Unmanned Aerial Vehicles) and TLS (terrestrial laser scanner) surveying systems will be integrated in a digital forest model at the service of planning, monitoring and optimization of forest operations.
SLOPE focuses on adding value to the mountain forest production through reliable stumpage price evaluation and quality control of the harvested material, thanks to the integrated use of real-time novel sensing technologies inside forest machineries to improve log/biomass segregation, have up-to-date forest inventories and refine stand growth and yield models. To achieve this, intelligent sensor systems will be integrated in the cable crane/processor head/truck machines to acquire wood properties and build quality prediction models. In addition, tracking systems will be used to trace the material, from the site throughout the whole supply chain.
The collection and integration of forest information like topography, geomorphology, cadastre, roads, climate, etc. with harvesting data like extraction distance and direction, stand density, silvicultural management and with material origin, quality and availability, in a unique online interactive system, is another priority of the project. A detailed forest model containing this information will be accessible and available in real-time to a number of operators, like: logistic operators, brokers, forest owners and sawmills to be used as simulation, short/long term planning and monitoring tool for the whole production chain.
The creation of this innovative workflow requires the achievement of a number of objectives, starting from the delivery of an integrated forest surveying system to cope with the frequent lack of precise and up-to-date information about a specific stand. The formalization of a model-based forest operational planning system is then fundamental for the correct exploitation of the forest stand data in tasks such as simulation and planning of manual and machines operations on the field. Another step is the integration of novel intelligent components in harvesting machines with a focus on the delivery of a multi-sensor model-based quality control system for mountain forest production. Tree and log quality, as well as machine operations should then be monitored through a specific and advanced enterprise resource planning (ERP) system. Integration and validation of the developed systems in real-life scenarios constitutes the final step to be achieved.
The typical scenario in which SLOPE tries to adds its innovations starts from the forest property. The owner who desires to manage the forest for productive use, needs to define a plan to efficiently and correctly exploit his property. These can be “forest management" for long term planning and “pre-operational”, created before the harvesting activity begins with the purpose to optimize the production benefit, selecting the trees that meet the market requirements. The harvesting plan is done by professional foresters (private or public) which generally will perform at least a basic inventory study, an analysis of age, species, terrain characteristics, natural boundaries, environmental characteristics, soil type and health, forest and public road network as well as any other relevant facility. Harvesting plans are generally related to single parcel properties where the harvesting operation will take place.
Once defined, the parcels close to the harvesting time are visited by a forester who examines the plot for marking the trees for a more detailed plants inventory. This operation will generate a more accurate estimation of the commercial value of the standing trees to be harvested, the so-called stumpage value. Tree marking is probably the most important silvicultural task, since the identification of which tree can be harvested and which is supposed to remain, allows to define the future development of the forest and its characteristics (economic, environment and soil protection, etc.) but will also influence the following steps of the harvesting process.
Current forest marking is a manual process performed using models based on tree diameter and height to estimate the timber volume and timber products. These models have a limited accuracy due to the fact that are created based on regional information and they might be created years before the harvesting activities. Additionally, visual evaluation of standing trees to be harvested, even if done by trained and expert operators is still a subjective task which can be prone to errors and biases. Current digital measurement techniques such as Terrestrial Laser Scanner (TLS), UAV and 3D modelling can increase the accuracy of the pre-harvest timber product estimations. The right estimation of the timber assets is very important for forest owner’s incomes, as usually based on the result of the inventory, using the stumpage value as a reference, the standing forest is sold in an auction. This might be missing, for instance when the forest owner is also undertaking harvesting operations.
Harvesting activities and log storage are usually performed by specialized companies entering the parcel for planning operations and asking for cutting permissions to the competent authority. This is the stage when the cable yarder lines and their hauling corridors are designed, together with landing area, storage pads and other logistic aspects like trucks or tractors turning points, more suitable machines, etc. All of this planning is done in person, walking in the forest and with the support of maps and compasses. GPS devices are becoming the standard for identifying the boundaries of the parcel and move throughout the forest to plan the cable lines, but they are subject to errors and wrong evaluations of working areas, can lead to significant delays, to cope with unforeseen situations during the actual harvesting days.
The harvesting operations begin when all the above tasks are successfully accomplished and the season/weather conditions allow safe operations without snow or heavy rain. Manual felling of marked trees is followed by their transport to the landing area with cable carriage, where the forest workers are instructed for the cut-to-length according to production plans and customer expectations, taking in consideration the individual properties and the presence of defects inside each log. The harvesting products include timber/logs and residues, which can be converted in fuel biomass. Both products must be promptly removed due to the limited storage area typically available at landing pads in mountain forests. A proper cut of each tree is fundamental for the maximization of logs value and meet the market demands but is not always performed in the correct way or taking in account tree characteristics (e.g. branchiness, diameter, stem taper, etc.).
Usually, residues are directly chipped and delivered to the end user. The log assortments are piled in intermediate storage areas, possibly close to public roads and accessible to timber-trucks. Volume, quality classes and overall value of the piled timber is calculated at the intermediate storage area by a professional forester, providing a detailed description of the assortments and their corresponding value and the stock may be then sold in further auctions. Even in this case, the evaluation relies on the experience of the forester and can be biased. If a sawmill has not directly bought the standing forest and contracted a forest company for the harvesting operations, the end user will contract a company for transportation to the facility deposit yard once purchased the desired assortments and volume. Here the logs are further sorted according to the requirements and the organization of the yard. This raw material will then be available as an operative stock for processing according to the industrial requirements of the end user. The stocking of logs on temporary storage areas can leave them under bad weather conditions that might affect their quality and value (e.g. parasites attacks, fungi, etc.) and must be reduced to the minimum.
The described workflow highlights a number of weaknesses impacting the forest production process addressed by the SLOPE main objective: to improve, harmonize and optimize the wood processing chain from the forest to the market. This can be achieved building an integrated process which fosters actor’s interaction and exchange of information, an integrated control system for real-time quality assessment and tracking, promoting sustainable forest production through transport and market optimization with a particular focus on mountain areas competitivity.
Project Results:
The SLOPE project studied and designed during the course of three years different technical and scientific solutions for the limitations affecting forest production in mountain areas. The key results can be summarized with a reference to the overall objectives of the project.
The first one consisted in the delivery of an integrated forest surveying system for mountain areas. SLOPE successfully investigated and tested on the field the combination of traditional surveys (forest plans, satellite imageries) with new sources of information (UAVs and Terrestrial Laser Scanning surveys). The acquired information combined to provide a multi-scale accurate description of the spatial and physical characteristics of the forest stand, resulted in a digital forest model with single tree level accuracy for quality (diameter at breast height, height, branchiness) and cutting instructions (distances between cuts for logs production).
The second objective, focused on the formalization of a model-based forest operational planning system for mountain areas. In this context, a final version of the SLOPE 3D harvesting planning tool has been developed and integrated inside the SLOPE portal. This last release introduced features like: rope launcher interface and cableway interactive placement configuration, visualization of detailed stem data and short-time cost estimation (felling, cableway setup, etc.), tree virtual marking, stand spatial queries, statistics and points of interest visualization, harvesting monitoring, routing and cost transport simulation, truck movements detection and point cloud visualization. The main innovation resides in the possibility to directly plan the operation on an online 3D graphical environment, simplifying the management of these complex operations especially on the mountainous environment where the third dimension becomes essential to correctly plan the activities.
The integration of novel intelligent systems in harvesting machines operating in mountain areas has been achieved through the creation or enhancement of ready to the market machines with features tailored to automatic log quality grading and traceability, like:
• Synthetic rope launcher to speed up the deployment of the main cable carriage line
• Intelligent cable carriage system with RFID antenna for log traceability, embedded computer for data recording and a Wi-Fi antenna for processor head communication
• Automatic chockers connected with the cable carriage for safe unlocking of logs after transport
• Enhanced processor head equipped with advanced sensors for automatic quality grading and RFID tagging
• Intelligent tracking system for logs transport based on an integrate board computer to be mounted on trucks
The aforementioned machines have been subject to scientific researches for the delivery of a multi-sensor model-based quality control system. More in details, SLOPE partners have defined a series of methodologies and indexes that can support the definition of the final quality-control for the wood production on the mountain areas. Four indices have been tested, based on NIR spectroscopy, hyperspectral imaging, stress-wave measurements and cutting forces. Each one calibrated to provide an output between 0 (bad quality) and 1 (excellent quality) for each processed log. The aggregation of these indices can provide an overall log-grading index. First promising results have been presented although further trials are needed to better calibrate the model. The use of robust RFID tags and readers across the whole production, certificates the origin and quality of each log to the final customer.
Another important aspect considered within the project has been the customization of existing enterprise resource planning (ERP) systems for mountain forest. In this case, the woodland manager platform has been improved with the integration of the 3D harvesting planning tool and the long-term planning IPTIM cloud service. The outcome, released on the market as Wuudis online marketplace, has been a comprehensive web portal with three main sections: marketplace, harvesting planner, and optimizer.
For the integration and validation in real-life scenarios, the most important project goal, the SLOPE consortium deployed the whole engineered system in two test fields under very different environmental conditions: summer in Italy and winter in Austria. The first one was fundamental to raise a number of system weaknesses that the consortium was able to solve for the second one. Overall, two weeks of testing were made, processing 27 trees and about 20 logs. The whole sensoring system collected data from eight logs, also used for comparing quality assessment with standard manual procedures. The rope launcher easer has been tested for one launch and the iTruck system was tested simulating two roundtrips from the forest to the sawmill. Cable carriage worked on the two pilots with only one day of interruption. Full wood traceability was achieved and preliminary test on automatic grading showed promising results although more tests on the field would have been advisable.
At the early stage of the SLOPE project, the consortium focused the identification of target users, understanding their requirements among the entire wood processing workflow and the prerequisites for the definition of the first integrated SLOPE system like the end-user interface, the system architecture, the common data model and the hardware to be installed/adapted/deployed during the project.
On a first instance the consortium identified the set of actors/stakeholders involved within the project and, after a description of the entire SLOPE scenario, defined specific questions targeted to each different kind of end-user. The answers were collected, aggregated and summarized in a set of guidelines to be taken into account during the course of the project.
After that, an analysis of the available hardware and software brought to the definition of the hardware required for the implementation of the SLOPE optimizations among the wood processing workflow going from the specification of existing hardware like UAV platforms or Terrestrial Laser Scanning systems to new hardware combinations like those adopted for the enhanced processors head or the solution for RFID tracking.
More in details, the sensors data transfer protocols and the compatibility with the control system of the processor head, lead the consortium to identify the CompactRIO of National Instruments as the development platform for both control system and sensors data acquisition and analysis, while the RFID system research on tree marking tags, field readers, mobile devices and hardware for intelligent trucks, found an appropriate and versatile alternative using Raspberry. A toolkit to develop devices prototypes integrating different signals and communication protocols for RFID reader, Bluetooth dongles, GPRS and GPS sensors.
The selection of the most suitable processor head after the evaluation of several roller and stroke-based machines, lead to the choice of the ARBRO 1000S. This machine features a stroke movement where the processor body holds firmly the tree, kept still, and the debranching arm moves forward. The absence of rollers enabled the processors to perform a first piling of timber to organize the stacks of different commercial classes and a lower productivity compared to roll processors was not influent, since the extraction of trees by cable crane was indeed relatively slow. In addition, the ARBRO processor performed the same activity (processing trees of the same diameter) requiring smaller prime movers compared to roll processors with lower fixed and variable costs as well as operative cost for the performed tests. Its relatively simple structure and electronics was particularly suitable for the envisioned modifications and integrations.
Regarding the cable carriage, the Tecno system manufactured by one of the consortium partners was selected for modifications while other forestry machines like the automatic chockers and the rope launcher easer have been designed and built from scratch.
As a follow-up of these requirements, the consortium defined the human machine interfaces for the SLOPE platform. The work has been devoted, on a first instance, at the analysis of the current state of the art, followed by the identification of involved actors and their functions. After that, a detailed analysis of the current end user interface has been reported as well as the definition of the final user interfaces for each device through a series of mock-ups for web, mobile and in-vehicle use cases. Adopting the principle of “least astonishment”, the current interaction metaphors have been revamped taking into account project requirements and consolidated workflows to introduce innovation without disrupting consolidated forestry actors’ behaviours.
Attention has been also devoted to the identification of all the available and derived datasets like trees properties, contextual information, harvesting data and wood quality, which could be useful inside the SLOPE forest data model for the achievement of an integrated and interconnected system. All these data have been listed and taken in account for the design of the central database.
The aforementioned actions paved the way for the design of the system architecture in terms of components, technologies and actual deployment. The design made large use of Unified Modelling Language (UML) Schemas to define the interaction between already available components and the missing ones, created following a federated approach. This architecture was not considered static but an evolving structure based on the challenges and the changing requirements encountered during the course of the project.
The requirement study was followed by the development of methodologies and tools to fully describe terrain and stand characteristics in order to plan, optimize and monitor, on-the-field operations. Starting from already existing hardware for low scale remote sensing and multispectral analysis, to reach detailed information on individual trees from high-resolution on-field surveys (UAVs and TLS), the consortium reconstructed and visualized the forest model in an interactive way.
After a first use of Rapid Eye Earth Observation satellite imageries for multispectral analysis where large areas of the terrain have been analysed under different indexes computations (NDVI, CCCI, NDRE, CHL) to detect vegetation type, density and health, aerial (UAV) and terrestrial (TLS) surveys were executed over two specific areas in Trentino, Italy, and on a second one in Annaberg am Lammertal, Austria. UAV flights above the forest using a fixed wing drone equipped with RGB and IR cameras were used to create a 3D dense point cloud (Dense Image Matching Algorithm) of the forest Digital Surface Model (DSM) and, in addition to other information such as Digital Terrain Model (DTM) or Digital Elevation Model (DEM), it has been used to derive a Digital Canopy Model (DCM). This survey was followed by a laser scanning of selected areas (plots) within the test sites, to detect tree properties such as branchiness and diameter and correlate them with the height of each tree. The combination of these statistics with position and height of each tree within a forest stand detected through RGB imagery segmentation and DSM, allowed the build of a virtual forest model where each single tree has its own precise properties and estimated cutting instructions. The work, considering a series of technical and operational limitations, has provided encouraging results.
The constructed model has been optimized for the visualization inside an interactive web 3D modelling for harvesting planning prototype. User requirements and envisioned human machine interfaces have been used to build an online three-dimensional virtual globe based on WebGL technology with high resolution survey data and advanced functionalities for data analysis, operational planning and real-time activity monitoring, such as: imagery and vector data visualization, DSM and DTM data streaming, point cloud visualization of each single plot, 3D forest visualization of trees and logs, measurements tools, cableway and rope launch simulation, spatial queries, virtual marking, cost estimation based on short-time optimization models, report generation and point of interest visualization. From a scientific point of view, high performance rendering of huge number of trees in real time and point cloud support directly inside common web browsers represents an important achievement in the exploitation of state of the art technologies at the service of forestry industry.
Another study carried on for the piloting activities has been the development of a logistic optimization model for a regional wood supply network based on an objective function for the minimization of transportation and other costs. This model supports the decision for the optimal supply network by selecting harvesting plots (origins), intermediate storages (terminals), forest roads and plants (destinations) for different type of products during a period in a specific region. To achieve this goal, analysis of all the relevant logistics locations within the forest for every pilot area, including skyline deployment sites, storage and processing areas, available roads and processing sites in the neighbourhood were performed. Part of this study has been integrated inside the harvesting planning tool.
Based on the acquired knowledge about actors, workflows and hardware requirements, a crucial result obtained during the course of the project has been the design, build and testing of machine prototypes to perform intelligent and smart operations in steep terrains. The following machines have been developed and tested on the field during the two pilot in Italy and Austria, subject to repairs and maintenance actions to improve their performances and cope with unexpected issues: a rope launcher easer, a set of automatic chokers, an intelligent cable carriage, the iTruck system and an enhanced processor head.
The chosen processor head was purchased and subject to a reverse engineering. A study of the mechanical 3D model and hydraulic schemes was then performed with the major effort in the modifications design of cutting forces evaluation, stress wave propagation, scan bar for cross-cut analysis and RFID tags positioning systems. Modifications involved both the processor head and the prime mover. In the first case, for the cutting forces evaluation, the built-in chainsaw was enhanced with linear encoders, flow meters and hydraulic pressure transmitters to detect transversal cutting energy and with load cells and hydraulic pressure transmitters to detect the blades debranching forces. The Near-InfraRed and hyperspectral measurements required the installation of a scan bar, consisting in a mechanical system housing sensors related to NIR, hyperspectral and stress wave (free vibrations) measurement systems. Each sensor was assembled on a movable bar parallel to the chainsaw, placed in a separated area for dirty, lubricating oil and chipping protection. Steppers, hydraulic cylinder, electro valve, inductive switches, protections, mechanical frames and components were used for the handling of NIR cameras, hyperspectral cameras and laser displacement sensors. The stress wave measurement as both time of flight across the log and free vibration measurement was possible thanks to dynamic load cells, hydraulic cylinders, electro valves and inductive switches at the service of mono and tri-axial accelerometers. Both sensors were mounted on two different movable parts, actuated by hydraulic cylinders. The cylinders push the sensors down on the log for the time of flight measurements and maintain the sensors in a safe position during all the other phases. The RFID marking system powering the processor head was built from scratch and mounted in a specially constructed carter in front of the machine. It was composed by a box kept in a safe position during machine operations and placed in front of the log (by means of a hydraulic cylinder) during marking activity. Inside the RFID marking box there was a hydraulic stapling system and an actuated RFID tag roll. Outside the box was placed the antenna devoted to reading each tag placed on the log. In the second case, the Liebherr R310B excavator has been equipped with an industrial PC, a touchscreen, an access point for wireless communications and a CompactRIO programmable controller, installed in the cabin. The work on the processor head has required a complete rewriting of all its basic controls extended with the new SLOPE functionalities. A specific human machine interface has been built for this purpose.
The design and build of the intelligent cable crane carriage on the basis of the technological solutions shaped by the consortium consisted in the installation of an external RFID antenna and a single board microcomputer with Wi-Fi connected with the machine communication system. Studies on the most suitable placement of the Wi-Fi antenna ended with its installation on the lower part of the cable carriage with a hydraulic mechanism to expose the antenna when reading and bringing it in a more secure position when not in use. Specific Data transfer schemes and protocols were defined for CAN Bus communication between the PLC controller and the single board computer and for the selection of in-process parameters to be transferred to the industrial PC installed on the excavator.
New types of automatic chokers interacting directly with the carriage for fast and unmanned detachment of logs after transport have been built and tested. They were equipped with an electromechanical element for moving the unblocking mechanism, an electronic receiver to decode the opening signal and a rechargeable battery. When lowering the load, the cable carriage checks the dynamometer and when the zero value is reached, it directly sends the open command to the chokers which automatically unhooks the load without the presence of an operator.
The rope launcher easer, completely built from scratch from one of the partners, was a system made to launch a synthetic rope from the beginning to the end of a cable line to facilitate the lay out of the main cable. It consists of a wheeled frame housing a power generator and a high-pressure air compressor. A slewing bearing in the chassis base supports the launch tube, which in turn is hinged to the tilt adjustment. Thanks to a GPS system, providing to the embedded computer the end coordinates of the line, the software calculates the launch parameters. The device is also equipped with an electro winch to collect the launched synthetic rope.
Studies on different RFID devices/tags models were performed on operational scenarios within the forest under different environmental conditions, to identify the most suitable ones for the project. Some trees samples have then been used for testing the dielectric properties of moist wood and its capacity to reduce RFID readability. Logs were gradually dried in oven, repeating RFID reading under the same conditions with different moisture content of logs, from maximum water content to oven dry. Since experiments confirmed their reliability, a new RFID UHF tag design was commissioned to the most suitable provider to be used for manual and automatic processor head tagging.
To bring intelligence to the transport vehicles for timber and biomass within the project scenario, providing real-time position, weight and fuel consumption for route and costs optimization, a hardware and software prototype called iTruck, was built. The prototype, made up of a single board computer, with GPS, Wi-Fi, 3G and Bluetooth connectivity, ensured recording and transmission of data to the central information system as well as logs traceability as it was able to connect with portable RFID readers used to scan the loads before transport.
A data storage back-up and management solution has been studied and built to handle data coming from on-the-field sensors, consisting in procedures and software running on the excavator industrial computer, to acquire, process and transmit data to the SLOPE central database.
In parallel with the development of the intelligent machines, a methodological basis for a semi-automated and real-time grading system for forest production was studied and applied on the processor head. Achieving this goal required an extensive testing and tuning of different sensors within the laboratory and on the field to ensure an optimal definition of thresholds and measurements for the final log grading. Grading is an important step in the production chain, when the resource properties and the quality of the product are assigned. Proper grading process leads to accurate definition of the timber value on the market, and in some cases, an extra value can be obtained for the timber products. Grading logs in forest is the first assessment of the technical quality of wood and it is a starting point for following actions in the fields of timber trade and its valorisation. Near-infrared and hyperspectral sensors have required more effort as the design of their mounting on the scan bar has been particularly complex. A proof of concept of a scan bar has been developed but its functioning was not tested on the field due to its fragility and the very harsh forest condition especially, on the Austrian forest site. However, promising results have been demonstrated in the lab using the scan bar on log slices taken directly from the field.
The SLOPE consortium after an evaluation of the state-of-the-art of the current practices and rules used for logs grading (user’s practices, legal norms, standards, regulations), defined five quality grading strategies to be tested, almost entirely developed using LabView software.
A 3D quality index, making use of laser scanners for 3D mapping of standing trees, was used to determine tree shape and stem dimensions (including diameters, taper, tree height, etc.). A custom software capitalizing on the partners’ experience was defined for this purpose and to simulate optimal cross-cut assuring the highest gain of the forest owner and compatibility with the local market demands. Results showed that basic log indicators could be easily measured using the stem 3D model created using TLS data with an estimate of the current indexes for each stand: waste ratio, products ratio, total profitability index and harvesting loss risk. Instructions and tree logs models saved on the central database were always accessible through the 3D planning and monitoring system.
A Near-InfraRed spectroscopy quality index was used for the determination of selected wood properties and defects. Chemo-metric algorithms for the prediction of intrinsic quality indicators were developed and validated using different types of sensors. Studies on the effect of wood moisture variation on the NIR spectra were performed on laboratory equipment and experimental data was used for extensive calibration. Experimental results highlighted the MicroNIR Spectrometer as the most accurate and suitable for the installation on the scan bar. NIR spectroscopy demonstrated to have a great potential for assisting quality grading of logs although there was no evidence of its application in the forest on the literature before SLOPE.
A hyperspectral imaging quality index was used for the characterization of bio-resources along the harvesting chain. As an extension of NIR over a whole area instead of a single line, this type of index was calculated through the development of chemometric grading models for qualitative analysis of spectra and a number of laboratory trials on wood disks taken from the field activities during the second pilot. Accordingly with the literature, hyperspectral imaging has a great potential for quality grading but its application within the forest was never been experimented. Trials lead to the design of a rig of 16 Hamamatsu sensors for the deployment on the scan bar.
Experiments validated these two approaches, as sensors were detecting the unique material properties of logs in objective and repetitive way, but the number of tested logs was not sufficient to assure precise quantitative analysis (such as direct computation of the chemical components content or detection of defects not previously included in the chemo-metric model).
The optimization of log/biomass analysis through stress-wave acoustic measurement was another evaluated measurement protocol. It was based on the observation of wave propagation and its correlation with mechanical and physical properties of the wood material (i.e. dynamic modulus of elasticity, density) as well as the presence of damage, decay or other specific defects. Although this approach was not new in the literature, the combination of multiple stress-wave measurements in a unique testing setup, the evaluation of different grading-dependent factors (e.g. spiral grain, knottiness) and integration with data from other sensors (e.g. estimated density values) constitutes a new research approach. Results showed that time of flight as well as free vibrations may be highly useful for quantification of log quality. However, the range of log characteristics, was still not sufficient to assure a precise quantitative analysis (such as direct computation of the value for modulus of elasticity or mechanical strength), but the quality indexes might still be applied for the screening of logs before shipping to the final user.
A cutting forces quality index has been defined with the analysis of two types of forces: cutting resistance of the chain saw cross-cutting the log, measured through hydraulic pressure and oil flow which are affected by the wood density and cross section geometry and resistance of the debarking knives during the stroke, measured by hydraulic pressure and mechanical deformations of the knife holder, which are affected by tree knottiness. These techniques, already used in secondary wood processing, represent a pioneering approach on the field. Results showed that measuring cutting power while cross cutting logs as well as when debranching, is highly useful for quantification of the quality and can provide additional resources not available in other systems.
All the aforementioned sensing techniques were used from the SLOPE consortium to develop a prediction model fusing each single index into a final quality value for the log/biomass that is being processed. Three approaches were tested: An expert system based on the “if-then” rules, multivariate analyses and statistical models (i.e. principal components analysis, partial least square regression, soft independent modelling class analogy and cluster analysis) and neural networks with learning supervised by an expert grader. After the tests, a linear weighted combination was finally selected as a data fusion algorithm, allowing customization of the properties importance for any quality sorting scenario through the setup of values for each quality index.
The development of forest machines and automatic quality grading strategies proceeded in parallel with the definition and development of the mountain forest information system (FIS), storing the data related to the harvesting process and interconnecting all the services available among the partners, creating a unique entry point for all the wood processing operations to the several involved actors (planners, logistic operators, brokers, end-users, forest owners). A comprehensive database was designed for these purposes, supported by remote web services to support the other SLOPE modules and clients for real-time control of the supply chain, short and long term planning as well as monitoring and selling of wood.
A number of software and hardware components working on top of the SLOPE FIS web services were developed for near real-time control of operations to provide planning information, storage optimizations, reduction of delays and advanced reporting taking into account critical conditions in the forest (network availability, rough environment, etc.). A first mobile app able to connect with handheld RFID reader and external GPS, was developed for field operations and database review, supporting tree RFID marking, tree/log status change for traceability, manual quality grading and forest stand analysis. The cable carriage internal software was modified to record tree RFID tags during the loading phase and record statistical data about the machine during its activity and transmit it to the black box for data backup on the excavator. The excavator and processor user interface and control software were re-engineered and rewritten in LabView to support prototype commands modifications and expose to the operator specific harvesting parameters. When the tree is collected with the processor head, the system shows its identifier, RFID tag, timestamp and suggested cutting instructions to produce logs while during the production of each log, the system shows length, diameter, volume, weight, branchiness and its quality index. A second mobile application has been also defined to create and manage storage areas and movements of logs assortments between them, while the 3D harvesting and planning system was updated to support monitoring of the status of a harvesting operation (storage areas, felled/processed trees statistics etc.).
The ecosystem for real-time control of operations has been the perfect case for additional studies of short-term optimization planning techniques. This process required an analysis of the best algorithms for cost and productivity calculations of several tasks involved in the wood processing chain, like transportation, felling, cable line setup, etc. minimizing storages, evaluating biodiversity and forest integrity impacts through a set of indicators (average slope, tree distances, height from terrain, surface area, obstacles, etc.). A selection of these models (cable way setup time, tree felling time, route transport costs) was implemented inside the 3D harvesting planning software to support forest planners in the estimation of times and costs of on the field operations. Values were computed in real-time based on selected terrain, surface and tree properties and could be exported as a report for on-the-field validation.
Medium to long-term optimization models were added to the SLOPE FIS for strategic and tactical planning of forest stands on a longer timeframe, providing instruments to optimize the use of available resources to satisfy the market demand of wood in the upcoming 5-10 years. This has been possible through the deployment of a software connector between the SLOPE FIS database and the IPTIM long term planning service. Forest owners can send their surveyed SLOPE stand data into the IPTIM service and receive a forest stand long-term management plan (i.e. felling, mechanical cleaning, soil preparation, pine planting, mechanical grass vegetation prevention, etc.).
The consortium integrated these long-term functionalities in a system for online purchasing and selling of wood products (timber and biomass), allowing identification and tracking of materials, negotiation, bidding, analysis and pre-selling procedures. The online portal called Wuudis is now public and available on the market as a full product, extensible with the SLOPE services once ready for the market.
To redefine the forest production workflow, surveys data, intelligent machines, software and hardware for real-time control and resource planning required the achievement of a complete integration of the different elements composing the project platform. For this purpose, the project envisioned an incremental integration plan having the FIS services and the forest common database model acting as the central communication hub between each platform component, with a web portal as the entry point for all the actors involved in the wood processing chain. A hybrid integration and validation approach, based on waterfall and agile-like development was executed. Three steps of integration with growing complexity have been done, each one with a sprint every two months and test cases validation followed by further developments.
The first integration phase, between the forest inventory and the harvesting systems focused on the access to the forest data model from all the hardware instruments, tracking systems and planning/visualization tools. Web services and the components accessing them were tested in depth to reach a percentage of satisfied use cases of 70%. The second integration phase, involved the forest management software. On-the-field data was connected with the central forest information database and used by the short/medium/long-term data analysis and synthesis services. Testing was executed to reach a threshold of 80% of completed use cases. Finally, the third phase devoted its effort for polishing and refinements to reach a test case completion percentage of 90%. Overall, a total of nine rounds of testing and validation were performed during the whole project to ensure the proper integration of the forest inventory with the harvesting system and the forest planning module. Satisfied test cases reached a percentage of 88%, a value not far from the target and affected by a missing set of functionalities foreseen at the beginning of the project for the ERP module but not implemented as already available inside other components. In parallel with the integration, each partner worked to refine and improve its hardware and software to reach a unified workflow. As an example, the 3D harvesting and planning tool added and improved the following functionalities: workspace save and loading, automatic generation of multi-language legend from maps, report creation and download, visualization of detailed stem data, tree virtual marking, spatial query and statistics visualization and point of interest visualization. All the software components on machines, mobile devices and fixed computers were able to communicate with the forest information system.
Three progressive techno-economic analyses of the SLOPE solution elaborated during integration and validation on the field highlighted its competitiveness with respect to the current techniques. Although a number of necessary assumptions, cost estimates shown a negligible difference of 1% between the two workflows approaches. The additional SLOPE surveys and processing costs were balanced with the removal of logs sorting costs, now directly performed during harvesting activities.
Two on the field pilot tests supported the aforementioned integration and validation activities with a major focus on the hardware of the intelligent machines. In a first instance methodologies were defined to evaluate the achievement of SLOPE during the pilot experiment, based on a detailed analysis of current systems by drawing process flow charts, identifying bottlenecks as well as possible strengths and risks of the supply chains. After that, responsible partners worked on the preparation of pilot areas.
The SLOPE project detected two valid tests sites in different period of the year and environmental conditions: north of Italy in summer and Austria in autumn. The activities started earlier than planned during the first year, with the UAV and Terrestrial Laser Scanning surveys on the Italian test site, while the other machines were tested only during the last year of the project. For the first demonstrator in Sover, Italy, the consortium marked trees with traditional and experimental techniques (RFID and tablet apps), identified the cable yarder landing area, the potential storage areas and concluded agreements with the forest owner and the forest/harvesting company for the operative aspects of the demo. Operational procedures, legal requirements (CE limits), safety instructions, felling the trees, installation of the cable yarder and carriage were studied before the actual deployment, updating time consumption, effort, productivity and costs. A similar process was carried out in Annaberg, Austria with a major complexity due to the logistics of bringing machines (whole system Tecno carriage, the excavator with the processor head, radio controlled chokers) located in Italy to the pilot.
The first pilot demonstration was performed in July 2016 for a total of one week of tests in parallel with regular harvesting activities. The SLOPE excavator processed a small number of trees while a roller processor head handled the majority of the assortments. The enhanced cable carriage worked continuously for the entire week with only one day of stop due to an issue in the power supply system for the single board PC. Collected and transmitted data were in line with the expectation. The processor head produced less data results than expected due to several issues encountered on the field like the overheating of the excavator or the required manual triggering of sensors, but was crucial to plan the improvements for the second demonstration.
Tests in Austria, in October 2016, ran more smoothly, even under bad weather conditions (snow), thanks to the automatization of sensor analyses and the revision of the excavator control system. All the planned 18 trees were extracted and processed in just two days and from them, eight logs were completely scanned and graded for quality. Overall, the processing speed was drastically improved and valuable experimental data for log quality grading was produced for further analyses. Preliminary grading results, in comparison with expert evaluations showed that the SLOPE system and the visual system are generally in good agreement, but due to the low number of processed logs the statistical confidence of the results was limited.
Compared to the visual subjective grading system, the SLOPE technology aims at being well-defined, objective and repetitive, trying to avoid subjective bias. However, reliable correlations between the sensor data, mechanical wood properties and its potential uses still needs to be developed to reach this goal. Well-calibrated sensors and models and the scientifically based definition of threshold values for the different parameters and their combinations are a pre-requisite for the quality rating. Once the system will be calibrated and the models established the openness of the SLOPE grading system will enable the definition of further categories of resource uses along the production chain by improved description of material properties.
On the field trials were fundamental for the improvement of intelligent machines, making them, in many cases, more than just a proof of concept. Full wood traceability was achieved and preliminary tests on automatic grading showed promising results besides more tests on the field would have been advisable.
The pilot activities were supported by a specific set of training material to be used on and off the field to transfer knowledge, skills and competences built during the project, to stakeholders and on-the-field workers. The consortium published a series of 15 video tutorials for the usage of the SLOPE online portal and performed demonstrations to forestry experts. The consortium planned also a set of communication tests for the cable carriage before the pilot deployment and published multi-language customer satisfaction questionnaires, several training meetings for the operators to explain the different work systems and prototypes, prior to the pilots. Additionally, the consortium organized and delivered a training course on the use of remote controls and load/unloading instructions with the Tecno cable carriage to the company in charge of the harvesting on the first pilot.
SLOPE performed a significant number of activities ensuring the dissemination of the described project goals to a wider community. This included publication of articles, organization of workshops, and presentations in conferences. Currently, the most important showcase for the project is its website (https://web.archive.org/web/20190818085734/http://www.slopeproject.eu/) constantly updated with all the relevant information, including the related events. A number of parallel web 2.0 communication channels have also been deployed and constantly updated with the aim of ensuring potential stakeholders’ involvement in the project. The consortium has been also active through a number of dissemination activities including scientific publications, meetings and organization of SLOPE-branded events as well as presentations of the project to national or international conferences.
The SLOPE project results and the acquired forward knowledge have been discussed and carefully analysed from the consortium after the second demonstration producing an exploitation and business plan and a shared licensing strategy.
Potential Impact:
SLOPE developed an innovative and cost-effective technology platform, providing a solution for planning, managing, measure and monitor wood extraction and production in mountainous areas. This is expected to have significant impacts in various areas.
From a scientific point of view, an integrated platform, a forest data model, novel technologies for forest data collection and reconstruction, 3D web visualization and planning tools and mechatronic improvements of harvesting machines will open new routes for the optimization of the wood supply chain.
From an economic point of view, the project will drive new products and services from partners to potential stakeholders interested in the exploitation of their foreground knowledge. Wuudis start-up and its marketplace or the rope launcher easer, soon on the market, are just some examples.
The current market demands raise a number of exploitable opportunities for the SLOPE technologies. Requests for sustainable forest planning and management is increasing within the global forest industry and forest managers need to find the balance between society's increasing forest products demands and societal benefits, and the preservation of forest health and diversity. Sustainable management of forest resources includes determining, in a tangible way, how to use these resources today to ensure benefits, health and productivity in the future. In this context, harvesting planning and optimization is an important foundation for sustainable forest management. This results in a growing demand for forest resources measurement and monitoring during the life cycle of the forest. However, the use of traditional methods has been proven not always cost-efficient for these new challenges. The main constraints for forest managers and landowners is the cost of the labour-intensive collection and management of the data and the activities related with the harvesting and extraction of the timber. The use of new techniques should address these issues.
A first opportunity for the SLOPE project is a better operational harvesting planning which is required in most European countries before proceeding with the harvesting activity. Operational plans are developed for each individual harvest area, based mainly on on-site inspections. Maps of the harvesting area, showing a detailed activity plan are required. Major points to be covered in an operational plan include the identification of:
• Harvesting area (usually 50-100 ha), location and boundaries (following topographic or natural features)
• Areas to be excluded from harvesting through prescriptions for flora and fauna protection, water quality protection or other identified reasons
• Silvicultural prescriptions to be adopted for different forest types
• Methods of tree marking for selection and protection
• Volume of wood to be removed by species and size classes
• Location, design, construction, maintenance and closure of roads, landings, log ponds and skid tracks to minimise disturbance to forest, soil and water resources
Current methods are based on a significant labour intensive work including field inventory, use of forest model analysis, technical GIS skills to create maps and harvesting mangers for the final decision making. Additionally, when using cable crane in mountain areas, engineering skills are required to plan the correct set up of the crane and other machineries in the field. Outputs of this work are usually documents with a set of maps including the location and design of the different infrastructure and silvicultural prescriptions. These outputs need to be shared within the different people involved in the project via email or paper maps.
Another opportunity consists in the provision of new tools to facilitate the different analysis by automating the forest inventory analysis, GIS analysis and planning of the cable crane as well as to combine these tools in a single platform (currently several software are used to process this information), enabling virtual collaboration of the different staff involved in the collection and analysis of the information required for the harvesting plan.
On top of integration, automation and sharing of harvesting plans, there is also an opportunity to improve the quality of the inventory data used as a base for the decision making. Currently it is mainly based on measurement of trees providing a rough estimation of their volume, based on models. The use of new techniques such as Terrestrial Laser Scanner (TLS), allows the creation of accurate 3D models of each single tree, providing a better estimation of the forest resources and facilitating the decision making.
These opportunities are supported by a clear trend of increasing harvesting operations. New regulations necessitate a more and better amount of information regarding harvesting plans in most countries around the world. As the technology evolves and social concerns about deforestation increase, harvesting plans need to be more detailed. The volume of timber harvested annually in Europe is 735 million m³ and 756 million m³ in North America, with a worldwide total of 3,359 million m³. Most of these harvesting operations are planed carefully from the start, representing a potential opportunity for better harvesting planning systems. A study based on FAO’s FRA 1990 dataset (updated to 1995 in 1997) estimated that there would be a considerable increase in the global round-wood supply from 1.800 billion m³ in 2000 to 2.275 billion m³ in 2040, which would largely be met by production from plantations. Although the productive forest plantation area was estimated at only 3 percent of the global forest area, in the year 2000, the study predicted that plantations would meet 35 percent of the global round-wood supply in that year, rising to 44% by 2020 and 46% by 2040. These figures shown a clear growing trend linked to even greater increase of the requirement for more detailed harvesting operation plans using new technologies.
A fourth opportunity resides in the timber tracking. New regulations (e.g. 2010 EU Timber Regulation) and certification schemas (FSC, PEFC) require that forest owners and managers ensure the wood was coming from where its suppliers claimed. Reducing illegal logging could substantially increase revenue from the legal trade in timber and halt the associated environmental degradation but law enforcement and timber traders themselves are hampered by the lack of available tools to verify timber legality. The use of new technologies such as RFID systems integrated in forest machines, transportation fleet and forest workers (chainsaw operator, forester), can offer a unique capacity to track and trace both trees and timber logs during the different stages of the supply chain. This technology will help forest owners and managers to meet their requirements of traceability regarding legal requirements and timber certification. At present, the main certification schemes, PEFC and FSC, cover a forested area of 236 Mha and 183 Mha respectively worldwide (in Europe the certified area is 95.5M ha and 81.8 Mha respectively). Therefore, there is a potential opportunity that a significant part of this certified forest will adopt automated traceably systems.
A fifth opportunity is related with the sales of timber. Wuudis Service, developed in parallel with SLOPE project, is a mobile and web service platform for real time forest management, care work and harvesting monitoring as well as online purchase and sales of timber, enabling detailed documentation and tracking of purchased wood, going from forest stand down to a compartment level. Wuudis is also equipped with features and functionalities for easy data sharing between all stakeholders in the value chain and for certification schemes.
The final opportunity is represented by the adoption of innovative machineries. In forestry, despite the increasing automation, is difficult to create a supply mechanization chain because forestry areas are always different and changing. SLOPE created a communication system among machines, an automated supply chain where man can rely on artificial intelligence and has a control task to perform plants evaluation, morphological definition, electronic marking of trees until harvest, weighting, calibration, the creation of a classified assortment and logistic transport. These machines have the potential to be competitive product for the market and some of them are planned to be included in the partners’ product portfolio of 2017. More specifically, the enhanced cable carriage will be equipped with a series of accessories such as the electronic scale, the tilt sensors, the optimization system of consumption and consequently pollution abatement, the back-up data system with Wi-Fi for the collection and storage of daily production data. The tag reading system with retractable antenna and automated radio control for the release of chockers will be offered as optional. The expected customers’ feedback would be the immediate idea of a carriage that, as well as having an excellent efficiency on the automatic work, provides an overview on the status of the forestry installation, on production rates and status of works. Automatic chockers are going to be launched on the market by mid-2017 after further development to make them lighter and easier to use while the rope launcher, the most welcomed product across the project is currently under certification (CE Mark) and is expected to be built at the end of the year or at the beginning of 2018. Its potential low selling price elicited enthusiasm and interest and many users’ operators have already contacted sales office with purchase requests.
From a dissemination point of view the work performed by the entire consortium has been devoted to the submission of scientific research articles participating to international conferences, project follow-ups proposals, submission of new dissemination contents to the official website and social media channels (YouTube, Facebook, Twitter, LinkedIn, SlideShare) and featuring on other online websites as well as keeping contacts with potential stakeholders.
Overall, 28 scientific papers have been published on national and international journals and conferences, with the consortium attending more than 30 conferences and more than 25 trade fairs. Four techno-scientific workshops and one final conference followed by an additional final workshop were organized. Results were disseminated with four press releases, five newsletters, two flyers, one infographic and more than 200 updates on social-media channels with 23 videos and 134 online slides presentations.
The project planned the involvement of industrial partners inside the consortium with the constitution of an advisory board (IAB) consisting of 3 senior experts, nominated and invited to assist to the demonstrations on the field. Two members out of three were in position to attend the first pilot and technical meeting in Italy, in July, while the third member managed to attend the second pilot in Austria. The board, after each meeting, provided valuable feedbacks on how to develop and further exploit the integrated SLOPE system results. An initial work was carried out to contact standardization technical committees (CEN/TC) and other public authorities to introduce them the first preliminary SLOPE results for quality grading.
In addition to the noteworthy amount of resources spent for the communication of results to potential stakeholders, the consortium worked on the definition of a business plan before the end of the project, using the exploitation booster service brought by the European Commission to FP7 and Horizon 2020 projects. For this purpose, a virtual meeting with business expert consultants was organized in December to define projects key exploitable results (KERs) and design their lean canvas models for future project evolutions. The outcome of this meeting has been a list of seven products and services with an accurate evaluation of innovative aspects and unique selling points, briefly introduced below.
Product/Service 1: Detailed digital model of forest viewable by many parties
Through the first service, large areas of forest can be surveyed in a small amount of time which is constantly reducing due to technology evolution. Aerial data capture is easier and more cost effective making it suitable for adoption and sharing with planners, foresters, and prospective buyers. With the innovative systems introduced by the SLOPE project forest owners can easily acquire more detailed data about their forest property in the inventory/planning phase. This service makes possible to generate a very detailed dataset with a cost-efficient system. UAV images are used for a first screening and characterization of the forest areas and, combined with in-field laser scans can generate a 3D forest model. The obtained model includes a highly-detailed canopy and terrain definition, as well as a full inventory of standing trees and the potential timber assortments in each single stem (bucking prevision). This result constitutes a feature unique in the market. Expected time to market is planned within six months.
Product/Service 2: Forest Planning Platform with 3D forest models
The second product is the real-time interactive 3D geographical visualization system. Using forest models acquired from UAV/TLS it helps in the planning, simulation and monitoring of forest production activities. Features include: 3D forest model visualization, open data, cable crane setup, slope analysis, measurements, working area setup, truck routing as well as spatial queries on the forest for timber product breakdown, cost analysis, cost forecasts and reporting. This service includes a completely web solution, globally scalable, supporting different forest actors with a level of interactivity unique on the current market. This can be used for a "virtual marking", identifying the trees to be felled according to their position, characteristics and commercial value. The model can be integrated with all the available digital data, such as property limits, road network (public, forest, etc.), protected areas, soil types, etc. Such detailed information, provided in an easy-to-use interface will return a reliable estimate of the stumpage value and harvest costs of the forest. It will also allow the owner to better understand the real value of its property, as well as facilitate the development of new forest management plans or update the existing ones. The model can also assist the chain of harvesting operations helping the definition of stumpage value either for selling the standing forest or for contracting the harvesting service, reflecting almost in real-time on the field activities (e.g. forester inspections). Expected time to market is planned within six months.
Product/Service 3: Automatic tracking and traceability system for timber products
The third solution consists in the integrated system for manual and automatic management and reading of RFID tags. In a market-adaptability focus, this system may provide the following services singularly or in a unique service:
• Tracking of single tree/log items and bulk goods along the supply chain
• Trace timber products with full details of location and date of key actions (felling, extraction, transformation, transportation)
• Transfer quality parameters and instructions for optimal handling/transformation by means of the same RFID system, maximizing added value (e.g. optimal value recovery)
This solution is the first auto-ID system integrated in forest machines, transportation fleets and forest workers combining widely available RFID technology with the devices and machines used during the raw timber supply chain, creating a unique system, accessible online and capable to track/trace both trees and timber logs. Most of the system is ready for the market but six month of further development and a partnership with a processor head manufacturer are needed to improve the solution robustness.
Product/Service 4: Wuudis mobile and web service
The fourth solution is an independent and neutral mobile and web service called Wuudis marketplace, including forest management, timber and biomass online sales and purchase service, and also a bidding service for forestry care works. This platform will cover also enterprise resource planning (ERP) for contractors in the next year. Currently, competitor’s solutions are closed services targeted only to their own customers/members. Wuudis is provided as free service for forest owners, requiring a commission fee to contractors or timber and biomass buyers only after a successful deal. Wuudis solution is already available on the market.
Product/Service 5: Sensorised processor head with wood classification grading
The sensorized processor head with wood classification grading is the fifth key exploitable result. An intelligent machine to provide automatic wood classification grading during tree processing, a unique feature in the market. The grading capability reduces uncertainties and errors in manual wood quality grading, providing real-time evaluation of wood under several different indices. Knowing in advance the wood quality, allows the owner to maximise its profits. Further developments and a partnership with a processor head manufacturer are needed to finalize the prototype.
Product/Service 6: Rope launcher for timber extraction
The sixth KER is consisting in the system to launch a synthetic rope from the beginning to the end of a cable line, to facilitate the placement of its main cable. The system helps reducing the number of hours required to setup a cable line as well as the physical effort for the forest operators increasing the productivity and the margins. Despite the relative technological simplicity of this machine, its concept is unique in the market. This is a brand-new product that can save days of work when compared with the current process of manually bring the heavy rope uphill. The product is ready for the market but is currently missing the proper certifications and is expected to be sold by the end of 2017.
Product/Service 7: Automatic chokers to unhook timber logs
The final exploitable result consisted in the automatic chokers to allow the operators to unhook the logs without the need of getting close to them, staying in a safe position. These automatic chokers present the advantage of directly interacting with the control board of the motorized cable carriage, directly sending the open command to the chokers only when there are no forces detected by the dynamometer on the cable, increasing productivity and safety. Expected time to market is planned within six months.
To understand how the vertical and horizontal products and services could work individually and, the consortium worked on a set of preliminary lean canvas models, to identify joint-exploitation plans, involved partners, key metrics, unique value proposition, communication channels, customers, early adopters, costs and revenue streams. Although the main focus was on individual exploitation of results, a SLOPE global platform as a services marketplace has been envisioned, acting as the potential integration space of the whole value chain, offering a unique shared economy model, linking in the future services such as Wuudis with other ones such as the 3D forest model, tracking system and intelligent processor head. Ownership was assigned to the different partners that may run one or several services based on the SLOPE platform while producers/suppliers were identified in forest managers, harvesting companies and forest public bodies. Finally, consumers creating the necessary demand, were identified as forest owners, harvesting companies, timber buyers and certification authorities.
Since the type of industrial research activities carried on by SLOPE could have an impact on future forest production, all the elements used and generated by the SLOPE project had to be protected by the appropriate IPR and licensing policies. Therefore, the consortium worked on an IPR document to identify each element used or generated by the project that can be under a license and to define which kind of license should be applied. Both hardware, software, data and documentation were considered and reported.
The SLOPE project did not constitute only a joint effort to improve the current forest production workflow, but also a mean for each involved partner to increase its competences and expertise to be more competitive on the market. This process went beyond the foreground knowledge acquired in forestry and was transferred within each partner’s individual field of expertise.
Partner GraphiTech, was able to develop new multi-purpose interactive GIS spatial functionalities from the 3D harvesting and planning tool, such as: measurements, planning, multi-resolution terrains, routing, OGC and Open Data support, spatial queries and point cloud visualization. The partner plans to transfer them to its commercial GeoBrowser3D (www.geobrowser3d.com) platform as a new set of features for the market for: advanced data management, terrain and custom layer visualization, planning and collaboration, point cloud and BinaryGLTF visualization, massive rendering of objects and OGC compliance. It also plans to further develop and exploit the 3D Harvesting Planning tool, advertising it through the GeoBrowser3D website, social network and contacts through its network targeting public administrations, urban planners, companies that deals with logistic of any kind of goods, Geo-ICT companies and tourism agencies listed inside the SmeSpire (http://www.smespire.eu) network database.
Partner CNR, was able to develop and test in-forest grading systems (processor prototype) for early definition of timber quality as well as develop and test an auto-ID timber traceability system based on RFID tags. As a research centre, it does not plan to go to the market, but to use the aforementioned results to further test the intelligent processor and approach the dedicated industry in search of a partnership and to work on a collaborative project proposal with PEFC in Ireland, Spain and Italy, involving forest owners and managers.
Partner Compolab finalized the development of an intelligent processor head optimized for working at cable crane unloading sites and able to perform a series of analyses on the processed timber, as well as marking each log with RFID tags reporting and storing the collected information. Besides some earthmoving machine operator license, Compolab acquired new engineering design skills for heavy machines, hydraulics, sensor electrical integration and conditioning, traceability systems, extending its network of contacts in the wood industry, harvesting manufacturers, forestry end-users and stakeholders. As an engineering office, Compolab plans to carefully evaluate the submission of patents in the field of automatic log marking systems and cutting forces estimation and to establish cooperation with harvester manufacturers providing their expertise for the integration of their developed sub-system on existing machines and the design of novel machine equipped with marking systems and/or grading capabilities.
Partner Coastway being relatively new to the use of UAV in forestry learned from SLOPE that there is a large market for their service but the technology is changing rapidly and they have to stay up-to-date with recent advancements. Additionally, pricing is strongly influenced by the covered area and flight authorization require time and effort to be obtained. Capitalizing on the experience of SLOPE, Coastway is moving into new markets in Scotland and Europe convincing landowners to use a combination of TLS & Aerial surveys, moving into fields beyond forestry such as environmental monitoring and coastal erosion. On this regard, Coastway has been invited to Horizon 2020 Project “Life”.
Partner MHG, during the project, developed a prototype for online purchase and sales market place which was extended to a commercial service under the Wuudis® trade mark, creating also a spin-off company for this purpose, called Wuudis Oy. Additionally, MHG build a commercially ready interface with Simosol/IPTIM for forest exploitation optimization and is involved in new project proposals with some consortium partners. During the project, MHG learned new IT technologies like AngularJS web framework, Ionic mobile framework, REST API based stateless application architecture and Jelastic: container based cloud environment. Its plans for the future are to pilot the launched Wuudis Service through the DataBio Horizon 2020 project in Finland, Czech, Belgium and Spain, searching sales partners in these countries, offer joint added-value services in partnership with Finnish forest planning service providers, apply for further Horizon 2020 SME programs and closing a first round of investment for Wuudis Oy.
Partner BOKU, during the project had the opportunity to test innovative machines and equipment in real operating conditions evaluating practicability, efficiency and acceptance and developing an approach for prototype evaluation, optimization models for wood logistics, quality control strategies via Near-InfraRed and Hyperspectral Imaging and collaborating with partners from other sectors than forest engineering. As a research centre, it is mainly interested in scientific results collaborating with other partners from research and industry for projects, publications, courses, training schools, workshops etc. For these reason, future plans include development of logistic tools, standards for evaluating machines and prototypes as well as technology transfer in steep terrain timber harvesting, usage of developed tools and models during student’s courses and work on new research projects.
Partner FlyBy was able to develop new software components for Earth observation data download and automatic processing for vegetation classification and continuous analysis through NDVI and chlorophyll indexes. Additionally, the partner deployed web services ready to supply geospatial products to the end users and was able to build contacts with new forest researchers and end-users. Plans for the future include the use of developed components as services for final users in forestry and agriculture (i.e. vineyards monitoring) as well as exploit the contacts achieved in the framework of the SLOPE project to develop new projects and new business.
Partner Greifenberg improved and created new forestry machines during the project that are almost ready for the market. Its plans are focused on bringing them to the market to speed up forest operations and security.
Partner Treemetrics defined through the course of the project a more integrated 3D forest model, with a new tree quality index, new use of the TLS 3D models for harvesting planning and new use of the Forest Warehouse 3D cutting instructions within the SLOPE system. Additional development was carried on for their mobile application including new functionalities like: RFID support, tree status update, better integration with remote systems, connection with SLOPE central database and integration of Treemetrics database with the one in SLOPE. Future plans include the provision of data to the SLOPE services, like the forest inventory data and the 3D model of the forest as well as support for on-the-field marking activities.
Finally, partner ITENE was able to build a new tracking solution using RFID technology, discovering a completely new industry sector without previous experience in the forestry industry. Future plans will be focused on continuing with the development of their traceability solutions within this new industry, making also use of collaborative funds like the PEFC Collaboration Fund together with some SLOPE partners.
List of Websites:
Website: https://web.archive.org/web/20190818085734/http://www.slopeproject.eu/
Showcase Video: https://youtu.be/Ml8nQzf17Mc
Social Channels:
Facebook: https://www.facebook.com/pages/SLOPE-Project/351131505025437
Twitter: https://twitter.com/SLOPEProject
SlideShare: http://www.slideshare.net/SLOPE_Project/presentations
YouTube: https://www.youtube.com/user/SlopeProject