H2020 GREENPATROL Project: Enhanced Galileo solution for pest detection and control in greenhouse fields with autonomous service robots
- Type Project
- Status Filled
- Execution 2017 -2020
- Assigned Budget 2.058.795,5 €
- Scope Europeo
- Main source of financing H2020
- Project website GREENPATROL
The main result of GreenPatrol is the implementation of a complete prototype robotic solution that implements the concept of Integrated Pest Management in crops, with the ability to navigate within greenhouses while autonomously detecting and treating pests.
GreenPatrol specifications were determined through interviews with potential users and stakeholders, as well as greenhouse analysis. The system architecture was designed from a functional and modular perspective. Tests were defined to verify compliance with each requirement at the subsystem level and for the entire system. Security mechanisms were also considered at both the hardware and software levels.
The GreenPatrol robotic prototype is the actual hardware-software system, which includes all the developments made in the GreenPatrol project. The GreenPatrol prototype consists of the following hardware components:
- Global Location Unit
- Robotic platform
- Robotic arm manipulator
- Sensors (encoders, inertial sensors, 3D lasers, cameras and lighting)
- Actuators (sprayer)
- Processing and communications unit
In addition to the GreenPatrol prototype itself, the project has also delivered a prototype global positioning unit, based on GNSS and enabled for Galileo, assisted by inertial sensors and odometry.
The accuracy levels of the absolute localization solution, combined with relative localization within the greenhouse and autonomous navigation capabilities, allow the robot to navigate accurately and reach target positions. This feature enables the system to carry out inspection and treatment missions according to the IPM strategy within the required frequency, ensuring that the GreenPatrol solution is a valid and efficient alternative to traditional pest inspection and control management.
The GreenPatrol concept allows for a significant reduction in production losses and chemical use, as well as an increase in quality and competitiveness, resulting in more sustainable agriculture, greater food security, and soil and water protection.
The project included a validation phase in which the prototype was tested in both simulated and real-world environments, achieving a TRL of 7. GreenPatrol will be commercially exploited according to a business plan that shows profitability starting in the third year.
The image database for pest detection and identification is a public outcome of the project. By completing a form on the GreenPatrol website, you can access the database and tags containing images of Tuta absoluta, Bemisia tabaci, and Trialeurodes vaporarium on leaves.
Outreach and communication activities included website updates, scientific articles and press releases, radio and television appearances, and other channels such as LinkedIn, Facebook, and Twitter, as well as participation in events, workshops, and conferences.
Three components of GreenPatrol can be operated separately: the navigation solution, the robotic IPM strategy, and the pest and disease identification system. Patents have been filed for the pest detection and identification module, combined with the handling capabilities, and for the IPM strategy.
The GreenPatrol solution focuses on addressing the EU's need to produce more with less, as the European population grows and the area available for agriculture gradually decreases due to increased forestry and urbanization. Greenhouse crop production (with a 22% increase in area since 2011) emerged as a way to protect crops from harsh weather, enabling year-round production. However, it also creates favorable conditions for numerous devastating pests and diseases that can cause losses of approximately 15%. This has significantly increased the need for pesticides. At the same time, legislative measures and requirements regarding the quality and safety of vegetables have become increasingly stringent. Consumer awareness has increased, and with it, the demand for pesticide-free products.
GreenPatrol aims to develop an innovative and efficient robotic solution for Integrated Pest Management in crops, capable of navigating within greenhouses while performing early, localized pest detection and autonomously applying treatments. GreenPatrol's navigation capabilities are based on novel Galileo signals and the implementation of sensor fusion techniques. These capabilities result in increased production quality, yield, and farmer income, while reducing pesticide use and their environmental impact.
The main overall objectives include: 1) A precise positioning robotic solution, capable of operating in the greenhouse, providing accurate and detailed pest maps for case-specific treatment decisions; 2) Integration of Galileo's best-in-class capabilities in low-light indoor environments; 3) Perception with visual sensing for online pest detection and reasoning mechanisms for efficient action selection; and 4) Control strategies for manipulation and movement planning based on pest monitoring feedback.
European agriculture faces numerous challenges, such as population growth, climate change, resource scarcity, and increasing competition. Furthermore, the land available for agriculture is shrinking, rural areas are becoming depopulated, and the agricultural population is aging. The current challenge is to produce more with less.
Greenhouses protect crops from harsh weather conditions, enabling year-round production. Integrated crop management approaches provide improved pest and disease control. However, the intensification of greenhouse crop production creates favorable conditions for devastating infestations that can cost up to 25% of potential revenue. Greenhouse crop production is growing worldwide, generating €46,377/ha across Europe.
GreenPatrol aims to develop an innovative and efficient robotic solution for Integrated Pest Management in crops, capable of navigating within greenhouses while autonomously detecting and controlling pests early. Key developments will focus on:
- Precise positioning to generate accurate and detailed pest maps in greenhouses (low-light indoor environments);
- Perception with visual sensors for online pest detection, including reasoning mechanisms for efficient action selection;
- Manipulation and movement planning strategies based on pest monitoring feedback.
Galileo's capabilities are considered a key technology for the proposed solution, as it provides improved performance in the face of signal degradation and multipath present in greenhouses.
GreenPatrol will achieve a significant reduction in production losses and chemical use, as well as increased quality and competitiveness, leading to more sustainable agriculture, greater food security, and soil and water protection.
The project includes a validation phase where the prototype will be tested in a real-world environment to achieve TRL 7. GreenPatrol will be commercially exploited. Its business plan shows profitability starting in the third year.
The main asset of the GreenPatrol project is the introduction of a robotic platform for autonomous and effective pest exploration and control, based on a decision support system for Integrated Pest Management (IPM). Unlike most robotic solutions for pest detection and treatment (mobile bases, including inspection systems), GreenPatrol features a manipulator arm on a mobile base that provides greater mobility to the inspection system and allows for analyzing plants in different situations and growth stages.
Smart and precise agriculture places significant demands on the development and implementation of new processes, which this project addresses. Artificial intelligence and an autonomous robot are used for the early detection and assessment of potential crop pests. A new IPM strategy was developed specifically for robotic agriculture, which is another objective. A large number of images are required to generate AI detectors, something that was not available prior to the project.
Currently, there is no automated method for directly monitoring greenhouse plants for pest detection using a robotic platform. The proposed automated identification will provide an objective measure for assessing pest status and potential spread. All data generated during pest detection is stored using a data management system that will be very useful for developing predictive models based on environmental conditions. Sharing the generated datasets will be highly beneficial to the scientific and agricultural communities.
Most greenhouse robots are specialized for specific crops and environments. Navigation typically involves heating rails or pipes, or complex environmental instrumentation such as beacons, tags, or landmarks. For autonomous robots, the precise topology of the greenhouse (number of rows, length, spacing, etc.) is often required. The GreenPatrol robot offers a fully autonomous solution capable of navigating without the aforementioned restrictions. It uses a combination of global and local information to generate maps of the environment and localize itself and detected pests. Global localization uses precise point positioning techniques for high accuracy at any location, and the Galileo E5's AltBOC signal to improve tracking and reduce multipath within the greenhouse environment.
- FUNDACION CENTRO TECNOLOGICO DE COMPONENTES (CTC)
Related projects
