Periodic Reporting for period 2 - ROBS4CROPS (Robots for protecting crops)
Période du rapport: 2022-01-01 au 2023-06-30
The most important challenge in European agriculture is the cost and scarcity of labour. Robots have replaced labour in several sectors of the economy. Agricultural robots have not yet reached widespread acceptance. The vision of ROBS4CROPS is an agricultural sector where robots will replace humans in all heavy and unpleasant work!
From a technical point of view, agricultural robots do not reach their potential because they are used as stand-alone units rather than as part of a complete, innovative robotic system. From a non-technical point of view, agricultural robots do not fit well with current farming practices and agricultural standards and are not supported by an ecosystem of stakeholders.
ROBS4CROPS will tackle technical challenges by creating a robotic farming solution that consists of three elements: smart implements, autonomous vehicles, and the farming controller. Existing agricultural implements and tractors will be upgraded so that they can function, together with existing agricultural robots, as parts of a robotic system. Development and testing will take place in real farming environments (real operating conditions), in four countries, in iterative cycles, and in close collaboration with stakeholders.
ROBS4CROPS will tackle non-technical challenges by using existing agricultural standards, by utilizing existing machinery (thus lowering the initial investment needed), and by addressing the lack of maintenance, insurance, financing and training. Compliance with regulations, robo-ethics and socio-economic impact will be explicitly addressed. Robotics offers an opportunity to develop novel business models. Building the ecosystem for agricultural robotics will take place iteratively, in parallel with technical development. The complete robotic system (technical and non-technical) will be demonstrated at scale in pilots in four European countries.
From a technical point of view, agricultural robots do not reach their potential because they are used as stand-alone units rather than as part of a complete, innovative robotic system. From a non-technical point of view, agricultural robots do not fit well with current farming practices and agricultural standards and are not supported by an ecosystem of stakeholders.
ROBS4CROPS will tackle technical challenges by creating a robotic farming solution that consists of three elements: smart implements, autonomous vehicles, and the farming controller. Existing agricultural implements and tractors will be upgraded so that they can function, together with existing agricultural robots, as parts of a robotic system. Development and testing will take place in real farming environments (real operating conditions), in four countries, in iterative cycles, and in close collaboration with stakeholders.
ROBS4CROPS will tackle non-technical challenges by using existing agricultural standards, by utilizing existing machinery (thus lowering the initial investment needed), and by addressing the lack of maintenance, insurance, financing and training. Compliance with regulations, robo-ethics and socio-economic impact will be explicitly addressed. Robotics offers an opportunity to develop novel business models. Building the ecosystem for agricultural robotics will take place iteratively, in parallel with technical development. The complete robotic system (technical and non-technical) will be demonstrated at scale in pilots in four European countries.
The project is progressing according to plan. In WP1 (Ecosystems building), co-design sessions with a variety of stakeholders were continued. In WP2 (Smart implements) the smart sprayers and weeders were completed, notably by installing sensors to monitor the operation of the weeders. ISOBUS TIM was (partially) implemented. In WP3 (Autonomous vehicles) retrofitting of two conventional tractors for autonomous operation was commenced and nearly completed. LIDAR-based navigation was implemented to overcome the lack of dependable RTK-GNSS signals in the pilots in Greece and Spain. In WP4 (Farming Controller) the framework for communication between vehicles, implements, and the ERP was further developed and extensively deployed in the field. Work on orchestration and optimization of resources has been taken to hand. In WP5 (Testing) component tests and integrated system tests have been conducted for all four robotic systems. In WP6 (Large-scale pilots), work in the large-scale pilots has been coordinated, ensuring that scientific measurements were made to assess the KPIs that were defined earlier, and coordinating between technical developers and stakeholders involved in the pilots. In WP7 (Socio-economics and ethics) the focus was on a social cost-benefit analysis of agricultural robotics for each of the large-scale pilots. This work has resulted in a number of presentations at conferences and a journal paper. WP8 (Exploitation, communication, and dissemination) re-designed the project website, conducted social media campaigns, and supported dissemination efforts by all partners. Work also continued on the development of a comprehensive project-wide IP strategy. One outcome is that open-source may be the best strategy to leverage some of the project outcomes. WP9 (Management) maintained contact with the European Commission, coordinated the project, organized project meetings, and ensured timely delivery of all reports. Overall the project is well-placed to tie up loose ends and demonstrate autonomous robotics in all four pilots during the last year of the project.
ROBS4CROPS will develop components and integrate them into a practical autonomous robotic system for crop protection.
Non-technical aspects:
Exploitation: The consortium will provide business models for launching ROBS4CROPS solutions along with existing infrastructure and technology that farmers already own. Robots-as-a-Service business models will be examined to identify novel ways for accelerating adoption of robotic technologies.
Ecosystem building: An innovation ecosystem is “the set of actors, activities, and artefacts, and the institutions and relations, that are important for the innovative performance of an actor or a population of actors. This project will make full use of the existing Digital Innovation Hub projects and reach out to a wide group of ecosystem actors and elicit from them their requirements and concerns.
Ethics issues: We will apply ethical theory to review the specific applications of agricultural robotics and situate the questions about smart farming within the wider ethical debate about increased deployment of AI and robotic systems to perform tasks traditionally undertaken by humans.
Socio-economic costs and benefits: ROBS4CROPS will assess costs and environmental net-benefits and present the results in a web-tool that enables key stakeholders to assess the economic impact of varied implementation levels. For each scenario, externalities, non-market cost and benefits from robotic systems in comparison with conventional systems will be assessed.
Large scale pilots: A realistic test ground will be established in each pilot country to compare conventional farm operations with the newly introduced ROBS4CROPS solution. The large-scale pilots of ROBS4CROPS will connect with many farmers and other stakeholders. .
Technical aspects:
The robotic farming solution consists of three components: smart implements, autonomous vehicles, and the farming controller
Smart implements: ROBS4CROPS will equip agricultural implements with novel sensor systems and will develop methodologies to assess the performed operations. The project will use ISOBUS (ISO 11783) for communication between vehicles and implements. The Task Controller function of ISOBUS will ensure that a prescription map intended for a conventional tractor can also be applied by a robot.
Autonomous vehicles: ROBS4CROBS will implement full autonomous behaviour for autonomous agricultural vehicles. This will be achieved by sensors, models and software that are implemented on the robot itself. Full autonomy also includes task-level autonomy, which includes (re)scheduling of tasks in response to goals and conditions, such as backing out of a crop row which turns out to be blocked by an obstacle, then continue work in the next row; or “remove weeds from this field before this evening”. This will be achieved by planning and scheduling algorithms which will make use of a Digital Twin of each vehicle.
Farming controller: ROBS4CROPS will offer a fully autonomous robotic system for crop protection by transferring state-of-the-art controlling and synchronisation systems deriving from industrial manufacturing to the agricultural sector. The Farming Controller will be designed and implemented for establishing seamless communication of all resources and sensors with a digital representation of the field, under a common framework. The implemented model will host simulation tools as well as an upper layer controlling module
Non-technical aspects:
Exploitation: The consortium will provide business models for launching ROBS4CROPS solutions along with existing infrastructure and technology that farmers already own. Robots-as-a-Service business models will be examined to identify novel ways for accelerating adoption of robotic technologies.
Ecosystem building: An innovation ecosystem is “the set of actors, activities, and artefacts, and the institutions and relations, that are important for the innovative performance of an actor or a population of actors. This project will make full use of the existing Digital Innovation Hub projects and reach out to a wide group of ecosystem actors and elicit from them their requirements and concerns.
Ethics issues: We will apply ethical theory to review the specific applications of agricultural robotics and situate the questions about smart farming within the wider ethical debate about increased deployment of AI and robotic systems to perform tasks traditionally undertaken by humans.
Socio-economic costs and benefits: ROBS4CROPS will assess costs and environmental net-benefits and present the results in a web-tool that enables key stakeholders to assess the economic impact of varied implementation levels. For each scenario, externalities, non-market cost and benefits from robotic systems in comparison with conventional systems will be assessed.
Large scale pilots: A realistic test ground will be established in each pilot country to compare conventional farm operations with the newly introduced ROBS4CROPS solution. The large-scale pilots of ROBS4CROPS will connect with many farmers and other stakeholders. .
Technical aspects:
The robotic farming solution consists of three components: smart implements, autonomous vehicles, and the farming controller
Smart implements: ROBS4CROPS will equip agricultural implements with novel sensor systems and will develop methodologies to assess the performed operations. The project will use ISOBUS (ISO 11783) for communication between vehicles and implements. The Task Controller function of ISOBUS will ensure that a prescription map intended for a conventional tractor can also be applied by a robot.
Autonomous vehicles: ROBS4CROBS will implement full autonomous behaviour for autonomous agricultural vehicles. This will be achieved by sensors, models and software that are implemented on the robot itself. Full autonomy also includes task-level autonomy, which includes (re)scheduling of tasks in response to goals and conditions, such as backing out of a crop row which turns out to be blocked by an obstacle, then continue work in the next row; or “remove weeds from this field before this evening”. This will be achieved by planning and scheduling algorithms which will make use of a Digital Twin of each vehicle.
Farming controller: ROBS4CROPS will offer a fully autonomous robotic system for crop protection by transferring state-of-the-art controlling and synchronisation systems deriving from industrial manufacturing to the agricultural sector. The Farming Controller will be designed and implemented for establishing seamless communication of all resources and sensors with a digital representation of the field, under a common framework. The implemented model will host simulation tools as well as an upper layer controlling module