Final Report Summary - E-FLYWATCH (Development of an innovative automated and wireless trap with warning and monitoring modules for integrated management of the Mediterranean (Ceratitis Capitata) & Olive (Dacus oleae) fruit flies)
Executive Summary:
e-FlyWatch trap is a unique, fully autonomous insect trap with integrated electronics and communication modules able to capture real time images of Mediterranean and Olive fruit flies and to transmit the images and other information such as environmental and location data. e-FlyWatch system consist of the trap and a centralised data collection service providing: real-time warning to end-users and historical analysis of infested areas presented in a user friendly web-application. The e-FlyWatch product enables the producers in the Fruit, Vegetable and Olive sectors to improve their production, limit the amount of insecticides/pesticides, reduce the labour cost for spraying activities and reduce the routine trap inspection. It is a powerful new tool that will certainly improve IPM and promote sustainability in the particular agricultural sectors.
Nowadays, Europe face major problems from the damages caused by medfly and Dacus. The e-FlyWatch system could become a powerful monitoring tool for European fruit, vegetable and olive growers and assist in detection, delimitation and control programmes. The e-flywatch system is based on the well-known McPhail trap used for early warning and thus easily acceptable while the main drawbacks of traditional techniques such as manual insect recognition, high labour costs, difficult to implement pest control measures and lack of monitoring the actual pest population density dynamics are overcomes using the e-FlyWatch technology
The main e-FlyWatch product features are: a) A novel trap based on the McPhail trap design which has proved successful and is used extensively by fruit and olives producers, b) An embedded camera module inside the trap that allows capturing of the insects in the trap, c) Advanced electronic components such as the open source embedded controller which is based on a modular architecture, trigger module responsible for the early detection of an insect when it enters the trap even if the insect size is very small, the sensors data storage unit, GPS antenna, temperature and humidity sensors and other peripheral components, d) 3G communication module with 3G/GPRS antenna for field use and transmission of real-time data to the central station, and e) Power module based on solar energy, responsible for the powering of the electronic subsystems.
The e-FlyWatch system trap is programmed to transmit in real-time data such as insect images, temperature, humidity, location coordinates etc. The images are analysed by specially designed software which uses advanced image recognition algorithms to identify the presence or not of Mediterranean and Olive fruit flies. The recognition is based on detecting various anatomy, pattern and colour characteristics that makes the type and sex (male/female) of an insect recognizable. Particularly, the developed algorithms used are the template comparison algorithm for the identification of an insect presence in a captured image, the Particle area measurement for the calculation of the relative area covered by the insect and the comparison of the insect size with the tolerance limits, the overall geometric matching algorithm which looks for positive identification on the shape of the calculation of the Mediterranean or Olive fruit flies and the geometric features matching algorithm for the identification of unique insect features, as for example the abdomen, wings and thorax.
Once the process is completed the results are forwarded to the central station where the developed web based software system with GIS functionality presents all data in a graphical form. The web-based software system incorporates monitoring of insect populations in a field, regional or national level, performs statistical analysis, allows user interactive communication and records the pesticides usage for each location.
The potential of the e-FlyWatch technology, as concluded by the SMEs and RTDs, is both scientifically, socially and commercially significant, unfolding a new era in research and innovation in precision agriculture since this is the first time that real time recognition of insects in McPhail Traps has been achieved successfully. In addition, with estimated return of investments of less than three years both for the manufacturers and end users, the e-FlyWatch technology will have potentials of high penetration in the world markets.
Project Context and Objectives:
e-FlyWatch trap is a unique, fully autonomous insect trap with integrated electronics and communication modules able to capture in real-time images of Mediterranean and Olive fruit flies when entering the trap and to transmit the images and other information (such as environmental data, location coordinates of the installed traps etc.) to a centralised data collection service providing: real-time warning to end-users and historical analysis of infested areas presented in a user friendly web-application. The e-FlyWatch product enables the producers in the Fruit, Vegetable and Olive sectors to improve their production, limit the amount of pesticides, reduce the labour cost for spraying activities and reduce the routine trap inspection. It is a powerful new tool that will certainly improve IPM and promote sustainability in the particular agricultural sectors.
Initially, a study of the main approaches employed by end-users with regards to monitoring, warning, mass-trapping, use of pesticides, Integrated Pest Management and Control techniques was carried out. Most of the effort has been dedicated on the detailed description and understanding the trapping techniques of the two insects in question including the study of the different types of traps utilized and the trapping procedures necessary for each methodology such as the trap density, placement, mapping, servicing and replacement. Finally an extensive fruit grower SME survey was performed to understand the end users practises and specific needs with regards to warning, monitoring, management and control of Medfly and Dacus.
Experiments were carried out to evaluate the feasibility and define the method for automated recognition and identification of Medfly and Dacus, through the processing of live images of insects entering the trap. Optical recognition algorithms have been developed for the purposes of the experiments using National Instruments Vision Software. Different lighting conditions were also tested. Experiments have been carried out in controlled insect incubation laboratory and in the field. Several trap designs were tested until successful results were obtained. The experimental trap has also been compared to conventional ones for determining whether the insects behave in the same way and moreover to validate the results of the e-FlyWatch trap. The results were satisfactory with more than 80% success in identifying the insects.
The examination of various technologies has been conducted with regards to a) transmission including low power wireless communication i.e. ZigBee, WLAN, GSM/GPRS for SMS messaging and satellite internet for remote areas with no access to GSM network and (b) the presentation of the Results involving GIS technology, web-based management software for monitoring, database for data storage, statistical analysis reporting, graphical presentation of Medfly and Dacus density field maps and IPM software for providing precise warnings for spraying with pesticides. Based on the state of the art, the specifications were prepared for both the selected transmission technology and results presentation features including all possible variations of the e-FlyWatch System. The specifications specifically address the trap design, the camera system, the wireless transmission, the presentation of results and management platform and the IPM software.
A prototype Optical Recognition Module (ORM) was designed and produced based on an FPGA- solution. The module included a camera board which integrated the triggering mechanism. The ORM was also here developed based on Scorpion software libraries. This first version of the software uses two main features to identify the target insects (size and wing pattern). For the prototype communication module a GSM/GPRS module was selected. In the case of the prototype power module a custom circuit was designed to charge the battery from the solar panel and to regulate the voltage. Two sensors were also added to the module (temperature, battery / solar panel voltage status). All prototype developments were also tested by the responsible RTDs. Even though the results were positive, during the field tests and despite the efforts of the RTDs for further improvement, the FPGA–based solution was abandoned due to specific weaknesses connected mainly to the optical recognition module.
A new Optical Recognition Module (ORM), Version 2, was designed and developed which integrates an Open Source Embedded Controller (OSEC) as well as a Camera, a Trigger, 3G communication and power supply modules. An Optical Recognition Software was also developed, which checks in real time for new images and information from the traps, applies advanced image processing algorithms for the insect identification and forwards these data to the central station database. The final ORM Version 2 developed was tested and its functionality and operation both in lab and field conditions proved to be excellent.
For the construction of the Trap different methodologies were examined, such as direct injection molding, RTV rubber molding and rapid prototyping using UV resin and a UV laser for the final design of e-FlyWatch prototype trap. The first method was chosen for the base, the top and the side covers of the trap, whereas the third method was selected for the camera housing. In total 10 prototype traps were constructed to evaluate performance. Testing of the version 1 trap included mechanical tests, waterproof tests and functionality field tests for the evaluation of the trap’s effectiveness. The mechanical behaviour tests proved satisfactory, ensuring reliability and long life span of the e-FlyWatch trap, free from any fracture or fatigue problems. IP waterproofing tests have also proved satisfactory. Field testing of the traps were done in orange and olive fields in Cyprus and comparable results to conventional traps were obtained.
In order to fit the new, version 2 ORM hardware inside the trap, a new prototype version 2 trap was designed and constructed. The mechanical, waterproofing, and weathering tests were repeated and have been proved very successful. The field test results were successful showing that the e-Flywatch Trap version 2 attracts insects by more than 90% in comparison with the conventional McPhail Traps.
A new communication protocol for e-Flywatch System version 2 that uses GPRS was developed. It includes the following features:
a) Start of transmission upon Trap & Local Station demand;
b) Packet count and acknowledgement of data sequence;
c) Confirmation of end of transmission; and;
d) Remote device configuration.
The protocol has been developed with high scalability, and, if needed, more functionality (such as sensor types, or new data related with it, etc.) can be added, remaining compatible without any further changes. Additionally, some mechanisms have been developed in order to avoid any information losses and to add remote configuration possibilities.
The development of the web platform has been build using robust and high performance technologies for web development such as Spring, Hibernate, CSS or JQuery. The final web platform allows the end user to manage all the traps deployed in the field and retrieve all the information collected in the field. All the information collected is presented in different ways, including interactive charts and GIS visualization which provides precise feedback in the status of the different areas of the field. Special controls have been created to control the GIS visualization, allowing the end user to personalize the presentation and retrieve the historical status of the field. Included in the main platform is the communication service, which stores the data received via GPRS by the deployed traps. This process is constantly waiting for new connections, after one trap connects; the main process creates a new thread to attend it in ordr to implement the communication protocol. This process is in charge of activating alarms when necessary.
During the development of the software a process of testing and validating the system was carried out in parallel, although it was in the last stage of the development when more exhaustive tests have been performed. The tests ensured the successful functionality of the final application as their objective was to find and correct as many bugs as possible to improve the integrity of the application.
The field testing was implemented in 2 phases. During the first phase the hardware and software of e-FlyWatch Trap version 1 were integrated. During the integration of all system components to the trap prototype as well as after its completion, the initial testing was performed. This included the verification of the correct communication between the developed modules and the trap functionality verification in laboratory conditions. Tests were performed for the communication between the Optical Recognition module and the Communications & Power module . At the end of the 1st phase the field validation was carried out using the Version 1 e-FlyWatch Trap. The validation aimed at verifying the trap from multiple viewpoints, such as the trap’s functionality with regards to its ability to capture the insects, the motion detection trigger functionality and the imaging system performance. As mentioned above the field results were not satisfactory, since the Version 1 e-FlyWatch Trap proved to have weaknesses. This resulted to the design and development of e-FlyWatch Trap Version 2 and for this reason two new tasks were added in order to upgrade the functionality and the system performance since the Version 1 could not be further improved.
During the second phase the new hardware systems and related software were integrated and validated. According to the new design all electronics are packed in a specially designed enclosure that is attached to the insect path inside the trap. The field tests of the e-FlyWatch trap system version 2 were performed in citrus and olive fields in Cyprus covering all summer period. The performance of the e-FlyWatch trap system version 2 proved to be very satisfactory without any functionality issues with continuous communication flow of clear images and data. The captured images and various information data (serial number of the trap, coordinates, the day and time of the event, max/min temperature, humidity) were sent, using a 3G network uninterruptedly, to the central station where the optical recognition software runs.
The demonstration of the e-FlyWatch Trap Systems version 2 involved field experiments in 3 selected citrus fields in Cyprus. For demonstration purposes, a workshop was organised, on the 20th of September 2013, in the conference room of Agriculture Research Institute (ARI) in Cyprus. During the workshop, the technology developed was presented to participants from private and public sectors. Most of them were agronomists, olive and citrus growers and governmental policy makers.
As for citrus case, within Task 7.2 the demonstration process involved demonstration experiments, in 2 selected olive fields in Cyprus covering the period from August to October 2013. The citrus and olive producers and the governmental policy makers also visited an olive field test site in Zygi village where the case of olives was analysed by the project participating partners. At the workshop, the payback periods of Cyprofresh and San Isidro were given as examples. The main comments from the participants concerned a) cost issues, as for example what will be the price of a single e-FlyWatch system and whether the e-FlyWatch traps will be sold as a package with preferential price and b) whether the system can be extended to identify other insects as for example cherry and berries fruit fly. Feedback was excellent since pricing schemes of e-Flywatch system discussed were acceptable from most of the participants – the end users of the technology.
Training material and demo videos were prepared. Through the instructions given in the training manuals and videos, users became aware of all issues (system installation, electronics architecture, software, optical recognition algorithms) related to the automated traps. An extensive patent search has been performed and relevant patents in different related topics to e-FlyWatch project have been identified. Based on this search, the SMEs of the consortium will submit a patent since unlike other systems, e-FlyWatch consists of novel features which are clearly eligible for a patent.
The project website, (www.e-flywatch.com/) has been created and was updated throughout the duration of the project. A plan for the Use and Dissemination of the Foreground, PUDF, has been developed. The e-FlyWatch technology was presented in the form of a poster in the 2nd International Symposium of TEAM (Tephritid Workers of Europe Africa and the Middle East). Also the technology was presented in the form of a publication in the ‘First International Conference on Remote Sensing and Geoinformation of Environment’ (RSCy 2013). Russell attended the ‘Fruit Logistica Exhibition’ that took place in Berlin, 8-10 February 2012. During the exhibition the Russell representatives have presented the developed technology and they have joined the discussions with the aim of gathering necessary information with regards to the marketing of the e-FlyWatch Technology.
A Business and Marketing plan was finally developed to ascertain the viability of placing the e-FlyWatch system on the market within 1 to 2 years from the project’s completion. The marketing plan included an analysis of the present situation and the competition, a SWAT analysis and a financial analysis including a Return On Investment from the point of view of the producers and the end users of the e-FlyWatch system. The Business and Marketing plan of the e-FlyWatch technology as concluded by the SMEs and RTDs is both scientifically, socially and commercially significant since this is the first time that real time recognition of insects in McPhail Traps has been achieved successfully. In addition, with estimated return of investments of less than three years both for the manufacturers and end users, the e-FlyWatch technology will have potentials of high penetration in the world markets.
Project Results:
WP1: Scientific Understanding of the e-FlyWatch System and identification of fruit and olive sectors needs
Task 1.1 - Medfly an Dacus Characterisation:
The objective of the task was to study the main approaches employed by end-users with regards to monitoring, warning, mass-trapping, use of pesticides, Integrated Pest Management and Control techniques. In addition, information has been collected with regards to medfly and dacus infestations, related damage of fruits and excessive use of pesticides. CNE has been responsible for the task and has carried out most of the related work, while input has been provided from the SMEs as well, especially from Cyprofresh and San Isidro, who are the end-users.
In more detail, after performing a detailed literature survey, and cross checking references, key characteristics of medfly and dacus (number of generations, type of damage, resistance, distribution, etc) have been collected and listed by CNE. Afterwards, the concept of IPM has been studied. Detailed information has also been collected on Monitoring, Control, and Trapping techniques for the two insects in question. Monitoring includes a variety of procedures used to observe, measure, and record over time the activities, growth, development, and abundance of organisms or the factors affecting them. Control on the other side includes all management techniques used to suppress or eradicate the medfly (C.capitata) and dacus (B.oleae). Most of the effort has been dedicated though on the detailed description and understanding of the trapping techniques of the two insects in question. The various traps have been studied in detail by CNE, who was also responsible for the actual design and development of the final e-FlyWatch trap. The study included not only the different types of traps utilized, but also the trapping procedures necessary for each methodology, such as the trap density, placement, mapping, servicing and replacement. The mass trapping methodology has also been included in the study. A comparison has also been made between the various techniques and each technique’s drawbacks have been pointed out, which is important in order to understand the reasons that lead to the e-FlyWatch idea and approach.
A Survey was performed to capture the end users practises and needs with regards to warning, monitoring, management and control of Medfly and Dacus. A questionnaire has been filled by SME end users in Cyprus (Cyprofresh-Sedigep) and Spain (Cooperativa San Isidro) and also by the e-FlyWatch expert entomologist from the Agricultural Research Institute of Cyprus. Itis important to note that the survey included several hundreds of fruit and olive growers, members of the two associations as well as government agronomist inspectors. The replies were very important, showing that even though practices are different, end users expectations are the same. The information was used also to identify and select the most suitable specifications of the e-FlyWatch System variants.
Task 1.2 - Optical Recognition of Medfly and Dacus:
The main objective of this task was to define by experimentation the method for automated recognition and identification of Medfly and Dacus, through optical recognition of real-time images of insects entering the trap. Experiments have been carried out in controlled laboratory environment (CNE has access to the insect incubation lab of the local Agricultural Research Institute) and in fields by modifying conventional traps used by fruit growers. CNE has been responsible for the task and has organised the lab and field experiments in Cyprus.
Initially, a modified McPhail trap has been tested at an insect incubation lab, and afterwards it was transferred to an orange and olive tree fields. The first step has been to design a new insect pathway that would be much narrower than the original one and that could integrate a camera for monitoring the insects as they enter the trap. The insects should be walking while monitored by the camera and not flying, otherwise much more expensive equipment would be necessary for the task (high frame-rate cameras). Using 3D CAD methods, a new device was specially designed to fit directly on McPhail traps sold currently in the market, in order to carry out the experiments. A camera and the necessary video recording equipment have also been appropriately developed and placed in the modified trap in order to monitor the modified insect pathway and record the insect activity.
Fig. 1 (see ANNEX - Figures) presents the last version of the experimental trap that has been placed in the fields in order to test its functionality in real conditions and further evaluate it. The final experimental trap has also been compared to conventional ones for determining whether the insects behave in the same way. The results were satisfactory with more than 80% success.
The experimental trap and the tests confirmed the effectiveness of the trap design and proved that an optical recognition algorithm can be developed for the automatic Medfly and Dacus recognition under real field conditions. It should be noted that the successful experimental results from the new modified design of the trap insect pathway is a breakthrough at global scale given the fact that no similar work at scientific or commercial level exists.
Task 1.3 - Specification of Wireless Transmission and Presentation of results:
The scope of T1.3 has been the examination of various technologies with regards to transmission and presentation of results. The transmission of the results from the traps took under consideration various options, such as: (a) low power wireless communication i.e. ZigBee (b) WLAN and (c) GSM/GPRS for SMS messaging and (d) satellite internet for remote areas with no access to GSM network. The Presentation/Reporting of the Results involved collection of data from large number of traps and presentation of results using GIS technology, graphical presentation of medfly and dacus density field maps, database for data storage, statistical analysis and information with regards to the spraying medfly and dacus pesticides.
In more details, a study of the state-of-the art in terms of wireless communications for WSN was carried out by ATEKNEA in order to identify the most appropriate solution for the e-FlyWatch system. Two wireless transmission approaches have been found viable, taking under consideration current practices followed by farmers and end-user SME needs. In the first approach, the traps are connected to the LS (Local Station) via a wireless link, and the LS communicate with the CS (Central Station) and the User Mobiles. These specifications require a mid-range wireless transmission protocol for the link between the traps and the LSs and a long-range wireless transmission protocol between the Central and Local Stations. As a second viable option, there’s the possibility of integrating the trap and the LS in a single module. In this case, every trap will have the capability of transmitting data to the CS.
For the first case scenario, different alternatives were analyzed such as Zigbee, WirelessHART and Enocean. Zigbee was selected as the best option for the Trap – LS link, due to its performance, cost, availability, and for being a protocol used for several years and applied to different sectors. Trial tests were carried out in order to test the performance of zigbee radio link working at different frequencies (2.4 GHz, and 868 MHz), and using different radio output power. The trials were implemented using hardware from Ember (2.4 GHz), and from Meshnetics (Atmel). For the LS – CS link, GPRS has been the only reasonable choice since the beginning. For the second case scenario the 3G/GPRS communication should be used.
ATEKNEA has also prepared the specifications for the presentation of the results. These include all the specifications in terms of how to present the data in terms of web-based management software including users and actions, software modules, Enhanced Entity Relationship (EER), Database, and GIS options. Finally, ATEKNEA and CNE have prepared the specifications for the IPM software with the help of agricultural SMEs (Cyprofresh and San Isidro) that will enable the user to inform the system about pesticide usage details (quantity, date/time and type).
Task 1.4 - End User Opinion for e-FlyWatch System:
The objective of the task was to perform a fruit grower SME end user opinion survey coordinated by the SMEs to identify and select the most suitable specifications of the e-FlyWatch Trap from the end user point of view and specific needs with regards to warning, monitoring, management and control of Medfly and Dacus. Cyprofresh was leading this task, while CNE has prepared the survey questionnaire. The questionnaire included also a section on current practices that has also been used for Task 1.1. This questionnaire has been filled by partners in Cyprus (Cyprofresh-Sedigep) and Spain (Cooperativa San Isidro), two large associations with more than 1000 members and also from entomologist experts of Agriculture research Institute of Cyprus, who support end-users all over Cyprus. The results were important to identify and select the most suitable specifications of the e-FlyWatch System variants.
Task 1.5 - Technical Requirements of the complete system:
Based on the information from the previous tasks a specification of requirements has been prepared including all possible variations of e-FlyWatch System. CNE has been the task leader of this task, while all the other partners have contributed. TI contributed in this task by providing their input concerning the requirements related to the vision system, both in terms of hardware and design.
The specifications specifically address:
o The trap design
o The camera system
o The wireless transmission
o The presentation of results and management platform
o The IPM software
WP2: Design and development of embedded system for optical recognition, communication and power modules
Task 2.1 - Conceptual design of Optical Recognition:
The process to find the appropriate camera has been a challenge since there are significant limitations to physical size, cost, type, data output format and power consumption for the cameras. Equally challenging was the selection of the motion detection/triggering and lighting conditions.
Regarding the camera, an intensive search performed by TI for suitable suppliers, resulted with a few candidates and a few cameras were purchased for testing. The C329 JPEG board camera was chosen for the conceptual design mainly because of the high flexibility when it comes to the optics and support possibilities. Regarding the triggering mechanism, TI worked on a trigger that is based on the fly obstructing a beam of IR light across the path into the trap. TI used two detectors so that the system knows the direction of the insect in the pathway. CNE developed a triggering mechanism using the two detectors as suggested by TI and tested it in the fields using the same experimental setup as in the WP1 field tests.
CNE performed field tests on artificial illumination using both IR and visible light, mounted next to the camera. It was found that visible light had the unwelcome side effect that it attracted many more species of insects, especially at dark. Visible light and backlight illumination were found unsuitable. Since visible light is impractical and ambient light is insufficient, it is not possible to use colour as a sorting criteria for the insects since IR light only gives Black & White (BW) images.
Task 2.2 - Conceptual design of Trap Communication & Power Modules:
ATEKNEA prepared a conceptual design of the trap communication module and a conceptual design of the trap power module. The trap communication module is responsible for sending the data to the Central Server, since it was decided to avoid the use of any gateway solutions, such as a local station for collecting data from nearby traps. A GSM/GPRS module was selected for this purpose.
Regarding the power module, the use of a lead-acid battery has been found to be the only reasonable option, following also the expert opinion of Pessl (SME partner). The charging of the battery will be made using a small solar panel and a custom designed charging circuit, in order to minimise cost. A conceptual design of the module has been prepared by ATEKNEA. Pessl has also provided information on the charging circuit, which they have been using extensively in their meteorological stations.
Task 2.3 - Optical Recognition Software Development:
After examining the software development alternatives, the consortium decided to proceed with Tordivel’s Scorpion option. TI undertook this development, with help from Tordivel in customizing Scorpion for the embedded hardware used in e-FlyWatch. Since Scorpion is meant for use in conventional computers, libraries and code developed with it should be migrated to the chosen FPGA platform. An application was first developed using the full Scorpion package. The key focus was on the performance in order to determine the highest possible accuracy of fly recognition. Following this work, the optimisation of the algorithms was made in order to reduce the amount of resources needed, still within the Scorpion environment (for PCs). TI and Tordivel worked together in developing this software, called “classifly”, using images extracted from the videos collected in the field by CNE.
Most of this work consisted in the development of image conditioning algorithms that transformed the picture in a way as to maximise the outcome of a series of gauging algorithms. A major focus was to remove unwanted information in the image, i.e. noise. Having singled out the fly from the background, a number of features are extracted and inserted into a feature vector describing the insect. Classification is then performed by comparing feature vectors to previously stored data. The two main features that were used to identify the target insect species are size and wing pattern. After having developed the “classifly” software for PCs, TI initiated the migration of these functions to the initially selected hardware platform (Altera FPGA). Finally, this migration was not concluded since the field tests showed that the FPGA solution was not capable of working properly in the field and thus a different approach has been developed by the consortium (see Task 2.7).
Task 2.4 - Design and Construction of Prototype Optical Recognition Module:
During this task, TI designed and constructed the first version of the e-FlyWatch trap Optical Recognition Module. This hardware solution was not the final one, since after the failure of field tests, a second version was developed (see Task 2.7). The Optical Recognition Module (ORM) consists of: a) the triggering and illumination mechanism, b) the camera and c) the main processing unit. Regarding the triggering mechanism, several options were investigated in the lab. The final solution developed by TI is a custom circuit which basically checks the signals coming from two photodiodes against a predefined threshold value. Initially a separate PCB was developed and tested for the triggering mechanism and afterwards this mechanism was integrated in the camera board.
Regarding the camera board, after investigating several digital and analogue camera options, TI selected Aptina’s MT9V034 chip and developed a custom PCB around it. The Aptina MT9V034 has a resolution of 752 x 480 pixels. The third main part of the ORM is the main system processor, responsible for processing the images and identifying the insects. After experimenting with various alternatives, it was decided to proceed with an FPGA-based approach. The DE2-115 kit from Terasic was chosen.
In order to be able to combine the above sub-systems and provide the necessary interfaces for connection to the rest of the modules as well, TI developed two interface boards, one for connecting the Power & Communication Module (developed by ATEKNEA) to the FPGA and one for the connection of the camera board again to the FPGA. The prototypes delivered are identified as e-FlyWatch version 1 which, after the field tests, led to the development of the final e-FlyWatch system, the version 2.
Task 2.5 - Design and Construction of Prototype Communication and Power Modules:
ATEKNEA undertook this task, which includes two separate developments. The first development is the Communication part and the second one is the Power circuitry.
Regarding the Communications circuit, a GSM/GPRS module with EDGE capabilities was selected (MC75i). The selected module includes also an external output signal to indicate its power status and an input signal to switch the module on/off. These signals are useful to control the GSM/GPRS module or to switch on the device only when needed, in order to reduce power consumption. The communication between this module and the Optical Recognition Module was also designed during this task. To send the data to the GSM/GPRS module, a serial data bus, based on the RS-422 protocol, has been implemented.
Regarding the power circuitry, a custom circuit was designed to charge the battery from the solar panel and to regulate the voltage. The circuit developed is based on a design from Pessl. The power module includes also the system battery and solar panel. Finally, two sensors were added to the Communications & Power module. The first one is a Temperature sensor, while the second one is a voltage sensor (ADC) used for monitoring the battery and the solar panel voltage status.
All the above were integrated on a custom PCB, designed and produced by ATEKNEA.
Task 2.6 - Testing of the Prototype Modules:
The prototypes of the Optical Recognition Module (ORM) and of the Communication & Power Module (CPM) have been tested by ATEKNEA and TI. CNE performed the integration tests with the trap designed.
Regarding the Communications part, ATEKNEA focused on testing the system functionality in various reception conditions and experimented with different antennas. An external GSM/GPRS antenna has been selected in order to achieve good performance. In case the GSM/GPRS signal is weak, it is possible to use antennas with greater gain, which will improve the reception. This is left as an option to the final system user. Regarding the testing of the Power circuitry, ATEKNEA focused on testing the system’s performance in terms of capacity to keep the battery charged. Tests were also performed on the field. The tests were successful and no redesign was necessary.
TI and CNE focused on testing the developed ORM. TI’s tests were initially oriented in the functionality of the Optical Recognition software (classifly). CNE provided TI with various videos of the target insects as well as other insects that entered the trap, taken from the preliminary tests. The software responded successfully in its PC version. As explained in Task 2.7 below, after the field tests in WP6, this software was replaced by the final e-FlyWatch recognition software that runs on a remote server. Regarding the testing of the hardware modules developed for the ORM, the Camera Module was tested by connecting it to a serial interface and running a terminal emulator. The FPGA system was also tested by connecting an external keyboard, mouse and display.
Task 2.7 Design and development of Optical Recognition subsystem and software Version 2:
The FPGA image acquisition solution performed well under control laboratory conditions, but, after the field tests were carried out and despite the efforts of the RTDs for further improvement, this solution proved to have weaknesses due to various issues connected mainly to the triggering mechanism, picture handling, saving routines and increased power demands due to multiple triggering.
For those reasons SMEs decided to assign to CNE the responsibility to prepare a second version of the optical recognition module. Considering all options, it was decided to develop the version 2 trap based on an Open Source Embedded Controller (OSEC). OSEC is an open source wireless sensor platform used for implementing autonomous and low consumption sensor nodes. OSEC consists of a PCB board with a built-in micro-processor and it has the ability to integrate a variety of supplementary PCB boards for enhanced sensor connectivity. On the e-flywatch project, CNE used a 3G/GPRS board for communication.
The electronics of e-FlyWatch version 2 trap consist of: a) OSEC, b) Camera module, responsible for capturing images of the insect when entering the trap and then forwarding the data to the Communication Module, c) 3G Communication module responsible for the data transmission, d) Trigger module responsible for the early detection of an insect when it enters the trap and e) Power supply module. Fig. 2 (see Annex - Figures) presents the electronics of e-FlyWatch version 2.
The e-FlyWatch version 2 trap is programmed to transmit once a day the images taken from the camera as well as the serial number of the trap, temperature and humidity data to the central server. CNE developed also the final version of the insect recognition software (see Annex Fig. 3) which is running on a dedicated server and in real time checks for new images and information (temperature, humidity, coordinates etc) from the traps. Once new data is received, the system processes the images and sends the insect identification results to the database. Image processing is based on advanced algorithms, such as an overall geometric matching algorithm for the calculation of the medfly and dacus insects’ shape, and a geometric features matching algorithm for the identification of unique insect features, such as the insects’ abdomen shape.
WP3: Design and development of e-FlyWatch Trap
Task 3.1 Conceptual design of Trap:
Several conceptual designs for the e-FlyWatch Trap were developed by CNE based on the laboratory and field experiments presented in Deliverable D1.2 as well as on the feedback from the end users and SMEs. The various designs are based on the modification of the McPhail trap, which includes a camera module for taking clear images of the flies as they enter the trap, as well as enclosures for the electrical and electronic parts. The main requirements for the conceptual designs were to:
a) Produce an efficient trap which could easily be assembled and disassembled.
b) Ensure that the results from a conventional and e-FlyWatch trap are similar.
c) Provide easy access to the electrical and electronic equipment.
The first conceptual design consists of 3 parts, the Top Cover, the Upper Base and the Bottom Base with the camera housing integrated in the Bottom Base. Three main design variants for the first conceptual design of e-FlyWatch trap were developed. The dimensions of the different variations are slightly increased in comparison to the conventional McPhail traps due to the electronic equipment that needs to be positioned internally between the Upper and Bottom Base.
The dimensions differences were made mainly to fit the new camera housing, improve the locking between the Top cover and the Upper Base for better waterproofness for the electronic parts and for the correct functionality of the Optical Recognition module.
The second conceptual design consists of 4 parts and the main differences from the first one are that: a) the camera housing is a separate part, b) the top cover is placed in the external area of the Upper Base for better sealing, c) the field of view and the working distance are higher than in concept 1. The proposed external geometry according to this conceptual design is similar to the conventional McPhail Trap with an empty space between the Bottom and the Upper Base for the placement of the electrical and electronic equipment needed. The disadvantage of the 2nd conceptual design as compared to the 1st one, is the increased tooling and manufacturing cost due to the 4th part needed.
Task 3.2 Design of prototype trap:
Taking into account the above-mentioned pros and cons, as well as the specifications of the electrical and electronic equipment needed, the design of the final e-FlyWatch trap was developed. CNE was the Task Leader. The prototype trap consists of 4 parts. The differences between the final design and the concepts that were previously presented mainly concern:
• The increase of the space for the electrical and electronic equipment
• The increase of the physical light entering the insect pathway.
• The easy assembly and disassembly of the electrical and electronic equipment, by opening the two side covers.
• The more efficient diffusing of the pheromone and liquid attractant odour outside the trap, since the base will consist only from one part and the odour will be diffused directly.
Two equal compartments were introduced to the base mirror to each other. The two compartments have openings for the placement of the electrical and electronic equipment without the need of the top cover opening. The access to the electronic equipment is much easier compared to the other concepts, without the need of opening the Top Cover of the e-FW Trap. For covering the equipment, two side covers were developed. On this design, the camera is placed in its own housing which in turn is mounted to the Base. The latter is joined with the Top Cover using the same locking mechanism as used in conventional McPhail Traps.
Three channels were introduced in the base for more efficient diffusing of the pheromone and liquid attractant odour outside the trap. By using these channels, the total aperture area is slightly lower than the one of the original McPhail trap but as the experimental field tests show this area is more than enough. The channel dimensions are such in order not to allow even the smallest flies to enter the trap through them.
The design of the Top Cover differs from the traditional ones since it is designed to be placed in the external area of the Upper Base. Thus the final e-FlyWatch Trap will be 100% waterproof and the electronic equipment will not have to bear any failure risk due to corrosion or increased humidity.
The camera housing is designed so as to be easily mounted and dismounted on the Trap Base. The camera’s housing in comparison to the previous conceptual designs allows better diffusing of light into the pathway, as it is exposed directly to daylight. The camera working distance and the visible length are both 20mm, which comply with the system specifications.
Task 3.3 Construction of functional prototype trap:
In this task initially the evaluation of different methodologies for constructing the prototype trap was carried out. The options considered are:
• Direct injection molding using CNC machined metallic molds.
• RTV rubber molding which is a pattern based process whereby the RTV silicone rubber mold is built from a finished SLA (Stereolithography) pattern. RTV rubber molding is an additive manufacturing process using a vat of liquid UV resin and a UV laser to build molds a layer at a time. After the construction of the molds the final material is filled in the void.
• Rapid prototyping using UV resin and a UV laser to build directly the Parts from resin.
For the prototypes, the chosen construction technologies were the RTV rubber molding for the Base, the two side covers and the Top Cover, while the rapid prototyping is chosen for the construction of the camera housing due to the fact that this part will be modified continuously, based on the experimental results. The direct injection molding is the most accurate method but also the most expensive one (in case of producing a small number of parts) and is used mostly in mass production. Also the use of rapid prototyping technology for the bulky parts would have been a costly choice, mainly due to the cost of the resin materials.
According to the selected construction techniques, for the prototype trap the material is bright yellow polypropylene with Pantone colour 012C (same as the traditional McPhail Traps) and for the Top Cover a weatherproof Plexiglas (Poly-methyl methacrylate, PMMA) which is a transparent thermoplastic, an easy to handle/process and low cost material (also used in various McPhail Traps). The wall thickness for all parts is 2mm, which is 0.5mm higher than the traditional McPhail Traps.
In total 10 prototype traps were constructed in order to be used for the testing of the trap itself, for the testing of its performance in capturing insects using different pathway shapes and attracting insects by adding various openings/holes for the dispersion of pheromone odour, and finally for the field validation of the complete trap system functionalities (insect identification, data transmission to the central station, presentation of results to the end-users).
Task 3.4 Testing of prototype trap:
The tests that were performed by CNE, within the framework of this task, and included mechanical tests of prototype traps, hardness tests, waterproof tests and functionality field tests for the evaluation of the trap effectiveness.
The mechanical tests were conducted based on the ISO 527 - Determination of tensile properties. The hardness tests (Shore D) were conducted according to ISO 7619. The results show that the PP material of the trap base compared with other commercial polypropylenes is harder (shore hardness d77). Moreover the hardness of the top cover material PMMA (shore hardness d84) is higher than the PP. For the PP the measured elongation is smaller (6,9%) when compared with commercial polypropylenes. On the other hand the tensile strength is higher (45.2 MPa) when compared with commercial polypropylenes. The PMMA values are in accordance with other commercial PMMAs (67.9 MPa, 7,6%, see Fig 3.3). As a conclusion, the performance of the examined materials proved to be very satisfactory, in terms of mechanical behaviour ensuring trap reliability and long life span of the e-FlyWatch trap, free from any fracture or fatigue problems.
Moreover IP waterproofing tests were conducted using a spray apparatus that sprayed water at a rate of 0.05kg/sec to the trap, thus simulating the rain water. The duration of the test was 4h and it was found out that the internal compartments of the trap are well sealed, since the humidity sensor limit values remained constant during the experiment. Also the optical observation showed that the sealing of the compartments is adequate.
The prototype traps were exposed in the fields for a series of tests. These tests aimed at thoroughly evaluating the trap effectiveness and functionality. The system electronics used for the field tests were the ones already placed on the field for the preliminary trap testing in WP1. The traps were tested in orange and olive fields in Cyprus. The experiments were made by alternating the use of the new traps with conventional McPhail traps, at predefined time intervals in order to be able to compare the results. Different insect path entrance modules were tested in combination with the e-FlyWatch trap. The results show that a very small entrance hole and the lack of transparency can create significant problems to the trap functionality. On the other hand a much bigger hole requires more advanced image recognition algorithms for the insect identification due to the fact that the insects can be appeared in the photos in different directions.
Task 3.5 Design and construction of prototype trap Version 2:
For the housing of the electronic subsystems version 2, the e-FlyWatch trap version 2 was designed and developed (Deliverable 3.2). All parts have been designed by CNE using 3D CAD tools. The version 2 trap mainly consists of the base, the cover, the insect pathway and the enclosure for the embedded electronics. The new electronic subsystems can be packed together in a very small assembly inside the trap, as originally planned. The only parts that are placed outside the trap are the photovoltaic panel, the battery and its charger.
After the construction, various tests were performed by CNE. The material test results were almost identical to the ones presented in Task 3.4 since the material and production techniques were identical as for version 1. With regards to the IP waterproofing test the results show that the trap is well sealed, since water was not traced after 4 hours spraying test with water quantity 0.05kg/sec. The weathering evaluation test results show that no visual defects or surface degradation was observed on the samples after exposure on weatherometer for 2300 hours which is equivalent to a total Irradiance of 7,31GJ/m2. Also the surface colour variation rate was 4 and the end view colour variation rate was between 4-5.
Moreover, the version 2 prototype traps were tested, by CNE, in the fields aiming to evaluate the trap effectiveness with regards to the attracting of insects, with and without pathway. Conventional traps were also used in order to compare whether the design is equally effective in trapping the insects. Also results from the Limassol Department of Ministry of Agriculture were collected. The version 2 traps were tested in fields of Cyprus during May and June aiming to evaluate the trap effectiveness with regards to the attracting of insects. The results were successful and very promising.
WP4: Design & Development of Local Station
Task 4.1 - Conceptual design of Local Station:
The local station has been designed as a module that can be attached on any trap. ATEKNEA was responsible for the conceptual design. In order to design and develop a Central Monitoring Station compatible with the Local Station, initial tests were carried out using the selected GSM/GPRS modem TC65i from Siemens, simulating communication data from the field. The modem TC65i includes JAVA embedded extremely facilitating the research tasks during the project, as well as the installation and the maintenance services that the system has to support. The modem also integrates the following Internet services: TCP, UPD, HTTP, FTP, SMTP, and POP3. The tests undertaken are explained in WP5 – Design & Development of Central Monitoring Station.
The conceptual design has also taken care of the data communication format. The GSM/GPRS link can send information wherever there is mobile coverage, supports the TCP/IP protocol and provides short message service (SMS) as well as encrypted channel communication. To determine if the GSM/GPRS technology has enough capacity to send all the information, the amount of data sent by the traps was calculated. A data example for an installation of 3 traps has been prepared by CNE, including temperature/humidity, number of insects and error log. The example was prepared based on the type and amount of data expected from each trap for typical system operation. It was supposed that three different log files are present and updated in the Local station: the Temperature/Humidity (T/RH) log, the Insect log and the Error log. The T/RH log contains periodic information on the temperature and humidity logged by the system, the Insect log contains data sent from the traps regarding the insects that have entered in each trap and the Error log is updated each time an error occurs.
Task 4.2 - Design of Local Station:
The power and processing modules of the local station as well as its housing have been designed in order to incorporate in the trap. Based on the specifications and the SME’s partners demands, the most appropriate GSM/GPRS radio module for the e-FlyWatch system is the MC75i from Cinterion, that works with EDGE protocol, instead of the TC65i. Additionally, this module presents low power consumption in power save mode with a competitive price and includes an integrated real time clock that could be used to synchronize the node wake-up.
The communication protocol is being developed in Java. After the local station establishes the connection with the central station, it sends the measurements acquired during a period of time through different messages. After each message reception the central station send an ACK response. Finally, central station sends a new configuration if it is required.
The protocol to communicate the local station with the central station has been designed. This design includes different types of messages based on the different information that local station needs to send. Each message, detailed next, includes a message descriptor field to identify the message type, and CRC field to avoid errors in the data decoding.
The messages defined to enable the communication protocol are the next ones:
o INSECT DATA: Message containing a fly detection, the probability of being female and the probability of being medfly.
o SENSOR DATA: Message containing temperature and humidity measurements.
o ERROR DATA: Error or misconfiguration detected in the local station.
o ACK: Reception acknowledgement.
o EOT: End of transmission.
o CONF: New local station configuration.
o CONF AKC: Reception acknowledgement for a configuration message.
Communication is enabled via TCP protocol through the GPRS module. TCP is preferred over UDP because is connected-oriented and is more suitable to implement the interchange of messages. Also some timeouts for the response reception are specified in order to improve the power consumption of local station.
Task 4.3 – Construction of functional local station prototype:
The functionalities of the Local Station are integrated in the Trap, together forming a new standalone device. This device includes all the hardware defined for the Local Station including the GSM/GPRS module, and the required sensors, and the hardware specified for the trap, the flash memory and the optical recognition module. As the Local Station hardware is now part of the Trap hardware design, a communication protocol is used in the link between the Trap & Local Station. Additionally, this communication protocol is used in the link between the Local Station and the Central Station, reducing the complexity of the system.
The communication protocol includes the following features:
o Start of transmission upon Trap & Local Station demand;
o Packet count and acknowledgement of data sequence;
o Confirmation of end of transmission; and;
o Remote device configuration.
The Protocol has been developed with high scalability, and, if needed, more functionalities (such as other sensor types, or new data related with it, etc.) can be added remaining compatible without any further changes. Additionally, some mechanisms have been developed in order to avoid any information losses and to add remote configuration possibilities.
Task 4.4 – Testing of Local Station prototype:
Using the GPRS communication module developed in the WP2 the communication between the Local Station and the Central Station was tested. In addition, using the central station server developed in WP5 the communication between the Trap and the Communication module used in the FPGA kit and the BeMicro SDK was tested. In the latter case, the interface used has been the RS-422 serial bus, which performed well and has been tested successfully for different distances between the trap and the communication module.
WP5: Design and development of Central Monitoring Station including IPM Software based on GIS end-user database
Task 5.1 – Design of e-FlyWatch Central Station:
The design of the e-FlyWatch Central station was performed by ATEKNEA. The design includes:
• User cases: Description of the use cases, that describe the actions that the user may perform on the system.
• Database: An E/R diagram has been designed, where different entities are identified and described.
• Software modules: Description of different software modules grouped in three classes, communication and data acquisition, data processing and user interaction.
• Technologies: This point describes the selection of technologies used for the development of the central station. The main technologies are described below:
- Spring: Spring Framework provides an alternative to the Enterprise JavaBeans, although is suitable to construct any kind of Java applications, it is mainly focussed in Web Applications.
- Hibernate: Hibernate framework simplifies the database connection and provides methods to easily implement the DAO layer.
- Web interface: Technologies as Jsp and Ajax are used to create the web interface.
- GIS Platform: Open source alternatives have been studied, but the most realistic proposal due to the necessities of the system, which require satellite photographs with a medium/high detail, is Google Maps. Google Maps provide a JavaScript API, that may be used to implement the required functionalities such as density maps.
• Communication protocol: A protocol to enable the communication between the central station and local traps has been designed.
• User interface: This point includes the design of the web interface and the navigation map.
Task 5.2 – Development of e-FlyWatch Central Station:
The development of the e-FlyWatch Central Station platform (see Annex Fig. 4) has been completed during the second year period, based on the requirements set by the consortium partners, and especially the SME participants. The requirements that generated during the first year of the project (Task 5.1) have served as a guideline for development. The web application has been build using robust and high performance technologies for web development such as Spring, Hibernate, CSS or JQuery. The final web platform allows the end user to manage all the traps deployed in the field and retrieve all the information collected in the field. All the information collected is presented in different ways, including interactive charts and GIS visualization which provides precise feedback in the status of the different areas of the field. Special controls have been created to control the GIS visualization, allowing the end user to personalize the presentation and retrieve the historical status of the field.
The main platform includes also the communication service, which is responsible for storing the data received via GPRS by the various field-deployed traps. This service is constantly waiting for new connections and as soon as a trap connects to it, the main service creates a new thread to attend the data that the trap will transmit. These new threads implement the communication protocol. The communication service is also in charge of activating alarms when necessary.
Task 5.3– Test & Validation of e-FlyWatch Central Station:
During the development of the software a process of testing and validating the system was carried out in parallel, although it was in the last stage of the development when more exhaustive tests have been performed. The tests ensure the correctness of the final application as their objective is to find and correct as many bugs as possible to improve the integrity of the application. The application testing has been done in a cyclic way, meaning that the functionalities are tested, the errors are corrected, and then the application is tested again.
WP6: System Integration and Validation of e-FlyWatch System
WP6 has been implemented in two phases. During the 1st phase the e-FlyWatch System version 1 was integrated and validated by CNE, ATEKNEA and TI. Because of the various unresolved issues during the extensive field testing of version 1, a version 2 was developed in order to improve the hardware and software performance. During the 2nd phase the e-FlyWatch System version 2 (designed and constructed within Tasks 2.7 & 3.5) was integrated and validated by CNE, with respect to system functionality (insect capturing, data transmission to the central station, presentation of results to the end-users) and performance in attracting insects.
Task 6.1 – Integration of Hardware and Software:
During the first phase of Task 6.1 the hardware and software of e-FlyWatch Trap version 1 were integrated by CNE, ATEKNEA and TI. The e-FlyWatch trap system version 1 is composed mainly by three modules:
(a) The trap itself with the camera housing;
(b) The Optical Recognition module and
(c) The Communications and Power module.
The first part of the integration concerned the integration and communication between the Optical recognition module and the Communication and Power module. Subsequently, all trap hardware components were connected to the trap in order to form the complete and autonomous e-FlyWatch system ready for field testing. In total, two complete trap systems have been produced and placed in the fields for validation purposes, prior to the final pilot demonstration. The Central Server has also been set up in order to continuously receive trap data and present them to the end-users. The e-FlyWatch trap system version 1 consists of the FPGA kit, the Communications and Power module, the regulator board, the solar PV charger, the battery, a timer and two interface boards placed in a box with dimensions 300 x 230 x 110mm which was mounted outside the trap. The e-FlyWatch system functionality is based on the operation of the Central server which is responsible for storing all trap data and presenting results through a graphical user interface to the end-users.
During the second phase of Task 6.1 new hardware systems and related software were integrated by CNE (e-FlyWatch System Trap version 2). The electronics were redesigned and constructed into a new e-FlyWatch System Version 2 (see Annex Fig. 2). The main hardware modules of e-FlyWatch version 2 trap are: a) OSEC, b) Camera module which is responsible for taking images and capturing the insect when entering the trap, and then forwards the image data to the communication module for further processing, c) 3G communication module responsible for the images and data transmission, d) Trigger module which is responsible for the early detection of an insect when it enters the trap and e) Power supply module. All electronics are packed in the enclosure inside the trap which complies with the initial system specifications. Flexible ribbon cables and FPC bus are used for the connections.
Task 6.2 – Initial Testing of e-FlyWatch System:
During the integration of the version 1 system components to the trap prototype, an initial testing was performed which included the verification of the correct communication between the developed modules and the trap functionality verification in laboratory conditions. The tests were performed by ATEKNEA and TI regarding the testing of the communication between the Optical Recognition module and the Communications & Power module, while CNE performed the testing of the complete system, with the help from the SMEs.
Initially, the communication between the Communications & Power (C&P) module and the FPGA board (Optical Recognition module - OR) was tested. The scope of the testing was the verification of the correct data transmission from one module to another. The data communication between the two modules is based on the RS-422 protocol. Several tests were performed in order to verify that all data were always successfully transmitted from the OR module to the C&P module. The tests were 100% successful, also due to the fact that there is a very short distance between the two modules and the design of the C&P module has included some 120 Ohm resistances to regenerate the RS-422 data signal in case of presence of interferences. In this stage, data communication from the C&P module to the Central server was also tested. The communication is based on cellular networks and it was functioning well even at relatively low network coverage. At very low cellular network coverage conditions, the communication may result impossible or very slow. This issue is though overcomed by saving the data and keep retrying as long as the data have been transferred successfully. It should be stated though that the traps are meant for functioning in areas where at least basic (GSM, GPRS) network coverage is available.
The initial testing phase was finalized by CNE which included the integration of all trap components and modules. The scope of these tests was to verify the complete trap functionality in laboratory conditions. The first tests of the integrated trap in the lab concerned the correct functionality of the battery charging mechanism and the correct functionality of the voltage regulator board that had to be used to power the OR module. An off-the-shelf solution was used for this issue and, in order to verify its functionality, the trap was placed in the outdoor testing area of CNE for testing the charger’s ability to keep the battery full. Regarding the correct functionality of the regulator board that was designed by CNE to act as the interface between the battery and the development board, this was tested in terms of stability of the voltage provided. No significant fluctuations were noticed during the testing period. Finally, the trap was tested in terms of functionality and capability to detect and identify objects (or insects) that were forced into the trap in a laboratory environment. A significant number of medfly insects were forced into the trap, in order to check the trigger functionality and the quality of the images captured by the OR module. Other objects (not insects) were also used in these laboratory tests. The laboratory results were good, in terms both of trigger functionality and image quality. Nevertheless, as it is explained in the next tasks, the field validation showed that the triggering mechanism needed to be changed in order to properly function in the field.
During the second phase of Task 6.2 using e-FlyWatch Trap version 2, the main issues that have been improved are the following: a) the developed trigger module was modified in order to activate the camera module even if very small insects passed through the insect path. b) the reflection problems due to the direct sun light were solved. c) The trigger sensors were modified appropriately in order not to be affected by the natural IR light. d) the response of the camera was improved significantly by altering the low level communication commands between the 3G communication module and the camera module.
Task 6.3 – Prototype Production of e-FlyWatch Systems:
Phase 1 included the:
o Development of 2 Optical Recognition prototype modules, by TI, which were tested in the fields.
o Development of 20 Communications and Power prototypes modules, by ATEKNEA, 2 of which were used for the field validation and several others were used during the laboratory experiments.
o Development of 10 version 1 prototype traps (including camera housing), by CNE, 2 of which were used for the field validation of the complete trap system functionalities (insect identification, data transmission to the central station, presentation of results to the end-users) and five of them were used for the validation of the trap itself and its performance in fly capturing. The rest of the trap prototypes were used during the laboratory experiments.
Phase 2 included the:
o Development of 5 complete prototype e-FlyWatch Trap Systems version 2, which were tested in the fields during the second phase of field validation and during the system demonstration.
o Development of 10 version 2 prototype traps (including camera housing), by CNE, 2 of which were used for the field validation of the complete trap system functionalities (insect identification, data transmission to the central station, presentation of results to the end-users), five of them were used for the validation of the trap itself and its performance in fly capturing and for demonstration purposes and the rest of the trap prototypes were used during the laboratory experiments.
Task 6.4 – Field Validation:
PHASE 1
The first phase includes field tests in two different locations. The first one is a field in Zygi and was used for about a month and then the traps were transferred to Dali for an additional month of testing. The map in Fig 6 shows the selected locations. CNE performed a series of tests in the field, using the Version 1 e-FlyWatch Trap for several months during summer, autumn and winter between 2011 and 2012. The validation aimed at verifying the trap from multiple viewpoints, such as the trap’s functionality with regards to its ability to capture the insects, the motion detection trigger functionality and the imaging system performance.
With regards to the insect capturing capability, the experiments were conducted by alternating the use of the new traps with conventional McPhail traps, in order to be able to compare the results. Different sizes and shapes of insect pathway modules were tested. The experiments showed that a very small entrance hole and the lack of transparency can create significant problems to the trap functionality. A much better option is the use of a transparent Perspex pathway with conical entrance and conical exit, in combination with a 12mm hole. Using this configuration the trap was able to capture the targeted pests with no problems. With regards to the imaging system the results show that the image quality has a good resolution for identification of the targeted insects. With regards to the motion trigger and the optical recognition module the field test results proved that the FPGA image acquisition has weaknesses due to various issues connected mainly with the triggering mechanism, picture handling, saving routines and increased power demands due to multiple triggering. A large amount of work and trials were carried out by the RTDs in order to improve the performance of FPGA image acquisition. Although the performance was improved, the parallel processing of capturing and performing image analysis proved to be too much to handle by the FPGA based system and thus it was purposeless to further continue with this solution.
PHASE 2
The second phase of Task 6.4 includes field tests in two different locations using the e-FlyWatch Trap Systems version 2 which splits the capturing and the recognition into separate subsystems. The field validation of the traps was made by CNE in two fields of Agriculture Research Institute of Cyprus. The validation process aimed at verifying the trap effectiveness with regards to the attracting of insects, the trap functionality and whether the images and data could be sent uninterruptedly from the trap to the central station where the optical recognition software runs.
Initially the version 2 prototype traps were tested, during May and June, in the fields aiming to evaluate the trap effectiveness with regards to the attracting of insects. The results were very satisfactory and are presented in Annex Fig. 7. The effectiveness of the captured flies was by > 90% in comparison with the conventional McPhail traps. That means every 10 insects that a conventional trap will attract, the new trap will attract at least 9. Bearing in mind that early warning for application of pesticides in most of the cases is 6-7 insects it can be concluded that the new trap is accurate by ± 1 insect or better.
Secondly the version 2 prototype traps were tested with regards to their electronics’ functionality. One citrus field and one olive field were selected. The experiments cover the period July and August. Annex Fig. 8 presents captured images from the installed traps during the field validation. The performance of the e-FlyWatch trap system version 2 proved to be very satisfactorily without any functionality issues. The captured images and various information data (serial number of the trap, coordinates, the day and time of the even, max/min temperature, humidity) were sent, using a 3G network uninterruptedly, to the central station where the optical recognition software runs. The software uses a custom application which is running in real time and checks for new images and information from the traps on specified FTP address. Once an image is received, the software processes the image using the developed image recognition algorithms and sends the result and the other information (min/max temperature, humidity, coordinates etc.) to the central station database in numeric format.
WP7: Demonstration of e-FlyWatch System
Task 7.1 – Demonstration of the system for the citrus case:
The demonstration of the system to the participating SMEs and to potential end-users was carried out after the successful field testing of the e-FlyWatch Trap Systems version 2, with continuous communication flow of valid clear images of targeted insects and other data.
The demonstration process involved field experiments, by CNE, in 3 selected citrus fields in Cyprus (see Annex Fig.9). The demonstration in the fields initially covered the period from August to September 2013 when the medflies population was low. It should be noted that the data from the demonstration traps were also collected during October 2013 when the activity of the insects was very high.
The demonstration process also involved visit to the test sites of Agriculture Research Institute in Zygi village nearby the Limassol town (see Annex Fig.11). The participants were citrus and olive growers as well as governmental policy makers. During the visit the project brochure was circulated. It should be mentioned that the distributed brochure was previously evaluated and accepted by SME participants as non-confidential. During the visit, the project consortium have joined an information exchange forum with the aim of gathering necessary information and suggestions for further improvement and enhancement of the e-FlyWatch Technology. Especially the experiences of the fruit growers and their opinion about the new technology were very valuable.
Finally, for demonstration purposes, a workshop was organised, on the 20th of September 2013, in the conference room of Agriculture Research Institute (ARI) in Cyprus. During the workshop, CNE with the contribution of ARI, presented the technology developed to the participants. People from private and public sectors participated in the workshop. Most of them were agronomists, olive and citrus growers and governmental policy makers. The project brochure was disseminated to the participants.
Task 7.2 – Demonstration of the system for the olive case:
As for citrus case, within Task 7.2 the demonstration process involved field experiments, by CNE, in 2 selected olive fields in Cyprus covering the period from August to October 2013. During August and September the dacus population was not high, however the activity of the dacus during October was very high. The results are presented in the e-FlyWatch Graphical User Interface http://188.121.62.146:8080/eflywatch using User name: pilot, Password: pilot.
The citrus and olive producers and the governmental policy makers also visited an olive field test site in Zygi village (Agriculture Research Institute field) where the case of olives was analysed by the project participating partners. The fruit growers express their satisfaction about the capabilities of the e-FlyWatch system and they were surprised about the response of the system, since the CNE research team arranged for the system to send direct information through SMS messages to some of the fruit producers’ mobile telephones for various triggering events. The fruit growers stressed the time needed nowadays using conventional traps and the fact that the spraying of pesticides based on partial monitoring using conventional McPhail and yellow sticky traps is time consuming. On the other hand the governmental policy makers were very satisfied since the new technology could have benefits to health and environment and to the economy as a whole by increasing the production and the quality. The inspection team of the ministry of agriculture expressed their interest for using the e-flywatch system specifically for remote locations with olive fields where accessibility is low.
At the workshop, examples of cost models for Cyprofresh and San Isidro were presented. The main comments from the participants concerns cost issues, as for example what will be the price of a single e-FlyWatch system and whether the e-FlyWatch traps will be sold as a package with preferential price and whether the system can be extended for other insects as for example cherry fruit fly and berries fruit fly. Feedback was excellent since pricing schemes of e-Flywatch system discussed were acceptable from most of the participants – the end users of the technology.
Potential Impact:
During the project implementation period, material has been produced that will allow the SMEs to disseminate further the project results to their extensive market network. e-FlyWatch is a complex project with multiple value streams. In the research WPs it has been made clear that e-FlyWatch will enable the end users to improve their production, reduce the routine trap inspection, reduce the labour cost for spraying activities and limit the amount of pesticides. Various developments within this period were important enough to make it clear that e-FlyWatch Trap System has a significant commercial value. The targeted markets of the new product are the Fruit and Vegetable (F&V) Sector and more particularly the segments of the medfly susceptible fruits and vegetables and the Olive Sector. However during the course of the project it has become apparent that the system can be extended to include markets for the control of pests beyond Medfly and Dacus. In particular pests like the cherry fruit fly, Drosophila, Suzuki, vinegar fly and apple fly could be targeted. In the following paragraphs, the dissemination activities and the multi-faceted market potentials are analysed.
Dissemination activities and exploitation of results
Dissemination activities were carried out in WP8 and are presented below task by task.
Task 8.1 – Training:
In the framework of Task 8.1 training material and demo videos were prepared. Through the instructions given in the training document the users are aware of all issues (system installation, electronics architecture, software, optical recognition algorithms) related with the automated traps. In particular the D8.1 involves:
• User instruction guide for the operation of the complete system and its sub-modules. A 19 step-by-step guide was created to describe the procedure how-to assemble the new trap. The first 8 steps (1-8) describe the way all the electronic components are integrated on the Open Source Embedded Controller (OSEC). Steps 9 - 12 show how the assembled electronics are packed on the environment proof enclosure box. Steps 13 – 15 show the way the enclosure box is mounted with the insect pathway and then on the e-FlyWatch trap. Steps 16 – 17 present the Power Module connection and how it connects with the trap (see Annex Fig.5). The last step (18) describes the verification procedure which the user must follow, to confirm that the trap is assembled correctly and ready to be installed in the field.
• User instruction guide for the field installation of the final e-flywatch trap version 2. The e-flywatch trap version 2 can be installed as easily as a conventional McPhail trap, a 4-Step guide is introduced that describes the way the trap is installed in the field. Firstly, the top is removed and pheromone is placed in the holding basket. Next the wire coming from the battery is plugged in and the device is switched on. The top is then restored to the initial position and the trap is hanged on the tree. Finally the PV Panel Stand is placed, facing South and the trap is ready to be used (see Annex Fig. 11).
• Programming Guide for uploading software updates in the trap’s electronic system. The Interactive Development Environment (IDE) is used for writing the code and uploading it to OSEC.
• Description of the software used for capturing and analysing in real time the images from the trap. The e-FlyWatch version 2 trap is programmed to transmit the captured images and other information like the serial number of the trap, the day and time of the event, temperature and humidity at preselected transmitting periods, not exceeding 24hours to an FTP server using 3G network. The Image Recognition Software runs uninterruptedly on the server and is responsible: (a) to check for new images and information from the traps in real time on a specified FTP address, (b) perform image processing based on the advanced developed algorithms in order to recognize whether the captured insect is an insect of interest, (c) to update the database of the central station with the number of the identified insects together with other information. Also, the results obtained in the field testing are presented in the e-FlyWatch Graphical User Interface (see Annex Fig.10) which runs at the Central Station. Graphical User Interface presents: a) the graphical results from the installed traps (captured insects, accumulated captured insects, temperature) and b) the density maps for different dates.
• Training Videos from CNE and ATEKNEA with emphasis on system integration and deployment. CNE has prepared a demo video (see Annex Fig.12) that shows the e-FlyWatch Trap System electronics version 2, the application with the image algorithms and the field installation of the trap. ATEKNEA has prepared a demo video that shows the e-FlyWatch Graphical User Interface.
Task 8.2 – Protection of IPR:
An extensive patent search, through www.freepatentsonline.com www.patents.com (US, WO, EPO, JP). has been performed and relevant patents in different related topics to e-FlyWatch project have been identified and included in deliverable D8.4. A list of the patents for the following broader areas have been tabulated in the deliverable. The areas that have been considered for possible patent application are:
o Design of the trap
o Insect lure methods
o Design of the automated electronic monitoring system
o Design of the insect path, trigger and recognition system
o Integrated pest management system
o Communication system
In conclusion, the uniqueness of the e-FlyWatch system lies in the design of the insect pathway, the use and position of multiple optical sensors for trigger mechanism which ensures a high triggering success rate and a picture chamber design producing the correct lighting for the camera for clear real-time image capturing. The e-FlyWatch central monitoring system could include several intelligent modules, such as statistical analysis of historical medfly and dacus records and prediction forecasting models, as well as pesticides usage records and evaluation, etc. Advanced analysis routines at the Central IPM Platform and image recognition algorithms can remain as a trade secret. It is concluded that the unique design of the pathway, lighting sensors and triggering module, as described in D3.2 and shown in Figure 4, are the novel features which are eligible for patent protection.
Task 8.3 – Dissemination of knowledge and market exploitation:
The project website, (www.e-flywatch.com/) has been created and was updated throughout the duration of the project.
Deliverable 8.4 is basically a draft of the Plan for the Use and Dissemination of the Foreground, PUDF, which is submitted as D8.5.
The plan included two basic sections:
a) The confidential section that describes exploitable results and related activities, and includes:
• A verifiable list of all intellectual property rights that could be applied for or registered
• A list of the results that may have commercial or industrial applications
• An outline of the rights each participant has for each one of the results
b) The public section, which describes the dissemination activities and summarizes:
• The ways that the participants are going to reach their target public
• The communication strategy
• A set of dissemination actions presented in a verifiable way to ensure that the EC can keep track of them
The discussion on key issues of the exploitation plan was based on the Consortium Agreement that was signed by the beneficiaries.
Deliverable D8.3 presents the work that has been carried out by all partners regarding the dissemination of the e-FlyWatch project results. Particularly the e-FlyWatch technology was presented in the form of a poster (in the 2nd International Symposium of TEAM (Tephritid Workers of Europe Africa and the Middle East) (see Annex Fig. 13). Also the technology was presented in the form of a publication in the ‘First International Conference on Remote Sensing and Geoinformation of Environment’ (RSCy 2013).
The full Proceedings of this RSCY 2013 conference are published in the SPIE Digital Library with the code name Proc. of SPIE Vol. 8795 87950X
(http://proceedings.spiedigitallibrary.org/proceeding.aspx?articleid=172490).
Russell IPM attended the ‘Fruit Logistica Exhibition’ that took place in Berlin, 8-10 February 2012. During the exhibition the Russell representatives have presented the developed technology and they have joined the discussions with the aim of gathering necessary information with regards to the marketing of the e-FlyWatch Technology.
Moreover the e-FlyWatch project results were presented in the exhibition EIMA (International Agricultural & Gardening Machinery Exhibition) that took place at Bologna, November 7-11 2012, through the participation of PESSL.
Leaflets were prepared and disseminated during the exhibitions and also a poster was printed for the needs of the exhibitions.
Task 8.4 – Business and Marketing Plan:
The purpose of the Business and Marketing plan developed under this task was to ascertain the viability of placing the e-FlyWatch trap on the market within 1 to 2 years from the project’s completion. Pessl, Tordivel and Russell are the three SMEs who will be marketing the trap. According to the consortium agreement each of the three SMEs have full and equal ownership of the results, while they have the right to exploit individually or bilaterally any or all of the main results.
The agricultural SME End Users, Cyprofresh and San Isidro have free Licensing rights and have the priority of first-hand knowledge of the e-FlyWatch system, installation and operation. They will also hold the right to purchase the end-product at preferential prices.
The Business and Marketing plan analysed the present situation in which the three SMEs have considerable experience and market penetration. Furthermore, the new trap will extend their international network, give access to new geographical markets in new sectors and consequently increase their revenues. The market is very large particularly in the citrus and olive sectors in the Mediterranean region as well as in other parts of the world. A SWOT analysis showed that the strengths and opportunities overshadow the weaknesses and threats. Competition is small at the moment mainly from institutes and universities. For example a system developed by the Taiwan Agricultural Research Institute that identifies the oriental fly is very expensive and does not identify the insects in real time. These systems have not reached yet the international market.
The Business and Marketing plan also developed the marketing strategy to be used. The financial analysis included in the marketing plan was applied to both the technical-manufacturing SMEs as well as the Agricultural-end user SMEs. In the first case the Return On Investment analysis showed that the investment can be recovered within a period of 2.5 years, whereas in the second case within a period of less than 3 years. The Business and Marketing plan concluded that the marketing of the e-FlyWatch trap is economically viable.
Market overview:
The EU is the world’s second largest producer of fruits, with 14% of the world’s production. Furthermore, the EU is a leading world producer, accounting for 80% and consuming 70% of the world’s olive oil. The MED-8 countries were formerly considered as the ones mainly affected by the medfly and dacus, but it should be noted that last decade’s climatic changes have increased the affected countries. The medfly is a major worldwide pest and affects all citrus fruits but also peaches, apricots, apples, pears, grapes, figs and many other fruits and vegetables. The olive fruit fly, Bactrocera oleae (Rossi) (formerly Dacus oleae), is a serious olive pest in most of the countries around the Mediterranean sea and the most destructive pest of olives worldwide.
The F&V production accounts for 3.1% of the EU budget and 17% of total EU agricultural production. F&V received 3.4% of total EU agricultural budget expenditures in 2005, a total of €1.8billion. In addition, the olive oil production for which aid was granted by the European Union is steadily increasing and will surpass 2.3 million tonnes for 2013/2014. Table 1 (Annex) shows the economic value of the main F&V that are susceptible to medfly, and thus the true dimension of the problem. The main EU citrus and olive production fields are given in Table 2.
It is noteworthy that fruit, vegetable and olive demand is steadily increasing. In 2005, more than 1 million tons of citrus fruit were imported, which corresponds to approximately 10% of EU production. This shows that there is still room for extra production. Olives and olive oil is important to the economies of many regions and demand is steadily increasing both in the EU and in third countries (http://ec.europa.eu/agriculture/markets/olive/reports/rep_en.pdf).
World production of olives is presented in Annex Fig. 14 and Annex Table 3 (http://faostat3.fao.org/faostat-gateway/go/to/browse/Q/QC/E). The main world areas of production are Europe, North Africa and the Middle East. Orange world production is presented in Annex Fig. 15 and Annex Table 4.
Environmental Impact
In Europe more than €7 billion is spent for pesticides (accounts for 30% of pesticide sold worldwide) while hundreds of millions of Euros of counterfeit pesticides are imported across Europe (2008, Crop Protection Association, Counterfeit Pesticides Across Europe - Facts, Consequences and Actions Needed). Despite tighter regulatory controls compared to other parts of the world, pesticides consumption in Europe is increasing, and thousands of farmers, their families and neighbours are irreversibly affected by pesticides.
Some countries like Switzerland and Germany have national guidelines for Integrated Pest Management (IPM) and actively promote it at the National level. Other countries have National pesticide reduction programmes (Denmark, Sweden and Norway) while others have experienced increase of 100% pesticide use in the last 10 years (e.g. Portugal) (NGOs demand stronger public and private commitment to the reduction of pesticide use in Europe http://www.pan-europe.info/Media/PR/060906.html). Table 5 lists pesticide cost in the Med-8 countries. An important variance in expenditures per hectare is observed. This can be attributed to the fact that PPP costs (Plant Protection Products) depend on type of chemical used, type and age of crop and number of applications per year needed. The average PPP expenditure is 89 €/ha. Table 6 shows the pestiside cost based on this average. Expressed in euros per ton produced, the average PPP expenditure is 50 €/ton.
One would expect that an investment of this magnitude tackles the problem. But this is not the case. In the absence of a monitoring and alarm system, farmers generally prevent infestation by over-spraying pesticides. The majority of this work is however in vain. The cost of the infestation by these insects is in the best case an inferior product, reducing the revenues by up to 50% (as in the case for superior extra virgin olive oil) and in the worst case the loss of a whole harvest (100% loss). Reports from Madeira (FAO/ IAEA-tecdoc-1475 study) show that due to the favourable climatic conditions, the loss of production due to Medfly lies in the order of 22 to 38%. With regard to olive fruit fly, the damage caused by tunnelling of larvae in the fruit results in about 30% loss of the olive crop in Mediterranean countries, and especially in Greece and Italy where large commercial production occurs (see Annex table 6.
More staggering are the cost figures associated with the measures taken to combat medfly in Spain, the largest producer of citrus fruits in the EU. The Valencia region, which produces about 70% of the fruit of Spain, has invested in the recent years around €9mil/yr to combat medfly (Tragsa, 2007). Thus it can be estimated that more than €30 millions are spent in the control of the insect only in Spain. Similarly, in Portugal, even with treatments, residual losses of fruit losses attributed to medfly are estimated to reach €6.6 million per year (Guerreiro et al.1998). Consequently, the on-time spraying can dramatically reduce the losses. Only in Madeira island, controlling medfly effectively can save about 2.2 million kg of fruit. This represents incremental revenue for producers valued at €1.6 million annually. In other Mediterranean countries, like Israel, Palestinian Territories, and Jordan, the annual fruit losses are estimated to be about €280million, which is about half of the total revenue produced by MedFly host fruits in these countries. Under the current control programs, the direct damage (yield loss & control costs) and indirect damage (environmental impact & market loss) amount to €150million/year. These global costs in control and losses can be extended to Italy, Greece, and part of France, where citrus, stone fruits and olives represent an important economic resource.
The key environmental benefits derive mostly from the reduction in pesticide use. Therefore, to a great extent, quantifying the environmental benefits of e-FlyWatch corresponds to defining the environmental costs of pesticides. As a result of improved IPM management, the SMEs - producers will minimise the use of pesticides, achieving better protection of the environment and human health. Health impacts that are related to pesticides exposure include many cases of illnesses per year and millions of lost work days. Pesticide exposure is associated with several types of cancers, serious psychiatric, cardiac, teratogenicity effects and eye defects (Organophospate insetisides, Pesticides News No.34 December 1996, p20-21), depression, cognitive impairment and chronic fatigue (R. Davies, G. Ahmed and T. Freer, Chronic exposure to organophosphates: background and clinical picture, Advances in Psychiatric Treatment, Vol. 6, pp 187-192 (2000)).
Furthermore, the use of the e-FlyWatch will serve to decrease all the negative impacts deriving from the excess use of pesticides, like: the contamination of surface waters (Claver, A., Ormad, P., Rodrı´guez, L., Ovelleiro, J.L. 2006. Study of the presence of pesticides in surface waters in the Ebro river basin (Spain). Chemosphere 64, 1437–1443); the decline of farmland bird species (Boatman, N.D. Brickle, N.W. Hart, J.D. Milsom, T.P. Morris, A.J. Murray, A.W.A. Murray, K.A. Robertson, P.A. 2004. Evidence for the indirect effects of pesticides on farmland birds. Ibis 146, 131–143); the excess use of water for spraying, which minimizes water reserves (Norm Input – Output data for the Main Crop and Livestock Enterprises of Cyprus, Agricultural Research Institute, Ministry of Agriculture, Natural Resources and Environment, Cyprus 2007); the decrease of biodiversity and resilience of the ecosystem, due to the destruction of natural enemies (see Madeira case Environmental benefits of Medfly Sterile Insect Technique in Madeira and their inclusion in a cost-benefit analysis, a Study sponsored by the Joint FAO/IAEA Division of Nuclear techniques in Food and Agriculture, 2005). Also, Medfly and dacus constitute main obstacles to organic fruit production. The use of the e-FlyWatch system to control the pest could increase the land area that can be used for organic farming production.
Apart from health and environment protection, this project will contribute to another social objective: the strengthening of employment. Firstly, the system can help avoid severe damage to fruit crops, which can lead to the abandonment of agriculture, with immense social and environmental consequences. Secondly, the production of high quality products (including those from organic agriculture) will promote the image of Europe as an environment-friendly tourist destination. Thirdly, excess use of chemicals and a crop disaster can damage tourism, as it depends on the conservation of natural and cultural heritage, which includes rural landscape and the rural life. Increased attractiveness of a destination and better F&V quality can result in an increased number of tourists. This, in turn, leads to increased consumption of F&Vs, with benefits for the fruit producing SMEs.
Socio-economic impact (Benefits to end users)
As insecticide costs account for a large portion of production costs, producer profit is low. For the year 2004 in Cyprus, approximately €1.5million was just the cost needed for medfly chemicals, for only 4590 hectares (316 €/ha) with oranges, grapefruit, mandarins and peaches. Taking into account that chemicals are most of the time over-used by farmers to avoid a crop disaster, the true cost of chemicals and corresponding labour cost is a lot more. The use of e-FlyWatch can reduce the cost of chemicals to the minimum needed. Furthermore, the financial costs of pest control operations are extremely high in certain areas due to difficult field conditions. More precisely, when the terrain is very steep, mechanization is impossible. Also, farms often consist of several small plots distant from each other and water sources are often distant from the plots. As a result, many man-hours are required to carry out control operations. e-FlyWatch system will also serve to decrease this labour cost.
Improved Competitiveness for F&V and Olive Producers due to the e-FlyWatch Investment
For the purpose of e-FlyWatch System, investigation was carried out in studies related to spatial population density of medfly (2006, A.Alemany et. al. “Changes in the spatial and temporal population density of the Mediterranean fruitfly (Diptera:Tephritidae) in a citrus orchard” Spanish Journal of Agricultural Research(2006) 4(2),161-166) and dacus, while interviews were held with representatives from the Agricultural Research Institution of Cyprus. Although only 1-2 traps/km2 is recommended for the purpose of medfly and dacus monitoring (for suppression) (Trapping Guidelines for fruit fly programmes, International Atomic Energy Agency, Vienna, 2003), it should be noted that this small amount of traps corresponds to very large homogenous fields, which is not the case in Cyprus, Greece, Spain and other Mediterranean countries, where the biggest amount of fruits, vegetables and olives is produced in small nuclei, due to the morphology of the ground and property status. Taking all data under careful consideration, it was concluded that an average of 5traps/ha (or 500traps/km2) are necessary in order to effectively monitor standard spraying or mass-trapping techniques as well as to meet a pan-European application of the system. From these 5 traps a proportion of male/female attractant can be used depending on standard practices. Also, the optimum positioning of the traps in the available space is of great importance.
Manufacturing Cost per e-FlyWatch trap is estimated to be just over 100 Euros, and the trap’s lifecycle to be at least 7 years. The investment involved with e-FlyWatch system will have a short payback period and is well suited for small scale producers that dominate this market. The cost is not prohibitive for individual fruit growers, thus increasing the adaptation rate of e-FlyWatch System and consequently the economic impact and potential. A Business and Marketing plan was developed to ascertain the viability of placing the e-FlyWatch trap on the market within 1 to 2 years from the project’s completion. The marketing plan included an analysis of the present situation and the competition, a SWAT analysis and a financial analysis including a Return On Investment from the point of view of the producers of the trap as well as from the point of view of the end users of the trap. The ROI analysis carried out for the cases of the two end users taking part in the project, one dealing with citrus fruit and the other with olive plantations showed a return on investment of 2.5 years. The Business and Marketing plan concluded that the marketing of the e-FlyWatch trap is economically viable.
An investigation was conducted to examine the possibility to apply for a patent. It was concluded that an application will be submitted in UK for a patent with claims focused on the pathway design, triggering and lighting concept which makes the e-flywatch trap unique in real-time identification of insects. The plan is to submit a patent within the first quarter of 2014.
The Technical SMEs with the assistance of the Project Coordinator have been discussing potential agreements for exploitation of the results. More specifically, the SMEs will proceed in 2014 in the production of the final trap design, the electronics hardware modules while an option is given to the SMEs to use the Scorpion insect identification software based on a licence fee agreement. SMEs agreed also in the market segmentation. The e-FlyWatch trap System is estimated to be released in the market in 2015.
List of Websites:
Project website: www.e-flywatch.com , e-FlyWatch GUI: 188.121.62.146:8080/eflywatch
• RTO Coordinator: CNE Technology Ltd
Address: Demokratias 5, Ergates Industrial Estate
POBox: 16104
Postal Code: 2086
City: Nicosia
Country: Cyprus
• Primary Coordinator: Dr. Panayiotis Philimis,
Email: p.philimis@cnetechnology.com
Tel: +357 22624090 Fax:+357 22624092
Web: www.cnetechnology.com
• Task Manager: Dr. Elias Psimolophitis,
Email: info@cnetechnology.com
Tel: +357 22624090 Fax:+357 22624092
Web:www.cnetechnology.com
