Services

We offer a specialised service focused on the acquisition and interpretation of image datasets captured through optical sensors. These datasets provide valuable agronomic indicators such as Leaf Area Index (LAI), biomass, plant count, and more. Key insights that reflect crop health and growth patterns could be generated by analysing these data. These image-based datasets are crucial for training algorithms that predict various agronomic parameters, such as plant growth, height, yield potential [...]

We offer a comprehensive testing service to validate the safety functions of autonomous mobile robots and robotic systems. Our tests use a high-precision laser tracker for accuracy to assess your robot's ability to detect and stop when encountering an obstacle, as defined in ISO 18497:2024 (Protocol included in part IV Annex D.2). These tests, conducted at our Montoldre site in France, will help ensure your robot's compliance with industry safety standards. A representative from your team must [...]

This service tests the performance of on-board detection systems in agricultural robots, focusing on their safety functions under challenging environmental conditions like fog and rain. Conducted at the PAVIN platform in Clermont-Ferrand, France, the tests evaluate sensor accuracy and response, identifying any false negatives or positives. A manufacturer's representative must be present to configure the system and ensure proper operation during testing.

This service evaluates the effectiveness of onboard geofencing safety systems in agricultural robots, ensuring they stay within designated work areas as required by ISO 18497:2024. The test involves the robot following various trajectories that challenge boundary limits. High-precision monitoring ensures compliance, with the robot required to enter a safe state if it breaches the boundary. A manufacturer's representative must be present to oversee the process.

The ARPA 4 test thoroughly evaluates the safety systems of agricultural robots, focusing on their ability to detect human-like obstacles and automatically stop to prevent accidents. It simulates real-world scenarios by positioning mannequins in front of the robot as it moves through artificial crop rows. Using a high-precision laser tracking system, the test measures the robot’s trajectory and speed at our Montoldre site in France. This ensures the robot can safely operate in environments like [...]

Accurate and robust positioning is crucial for the safe and efficient operation of agricultural robots! Before integrating a positioning system, manufacturers require reference measurements to develop or select the appropriate system for their applications. Our dedicated service evaluates positioning systems through rigorous test procedures. This service allows you to compare metrics such as absolute position in the terrestrial reference frame, relative position in each reference frame, speed [...]

Our service is designed to assist clients in identifying and evaluating potential agricultural applications for AI or robotics technologies originally developed for other sectors. Clients can provide detailed information about their technology, including its current applications and technical specifications. Our expert team will conduct extensive research to identify possible use cases related to the agricultural and food production sector, analysing factors such as feasibility, potential [...]

This service includes all essential actions to gather knowledge, design, and document the client’s needs regarding Testing and Experimentation Facility (TEF) services. Through close collaboration with the client, we ensure that all technical, operational, and methodological aspects are clearly outlined, forming a solid foundation for future testing phases. By conducting these preparatory activities, we deliver a comprehensive setup analysis that serves as the entry point for subsequent testing [...]

Our service provides access to grain samples and comprehensive technological quality analysis conducted at Arvalis facilities, enabling the development of AI-powered grain analysis solutions. Clients can work with well-documented samples from one or multiple species, enriched with detailed metadata such as variety, harvest year, and collection location crucial for training and validating AI models. Beyond sample selection and preparation, our experts assist in choosing, testing, and validating [...]

As part of this service, we test and measure, among other things, mechanical, physical, acoustic, radio, electromagnetic compatibility and electrical parameters for the purpose of assessing the overall user safety of agricultural, horticultural, forestry and food machinery, equipment and components. Verification is based on the requirements of standards and directives declared by the manufacturer. We carry out an initial assessment based on the design documentation provided or/and on [...]

When you’re developing an AI-based solution, you need to assess its accuracy and its reliability. Whether it is for detection or prevision purposes, we can compare the outcomes of AI with ground truth and measure the accuracy of the detection or prevision of an AI-based model. If the model needs improvements, we can provide new datasets; see S00102.

You can provide your own dataset, without labels, or ask us to build the dataset and label it so you can train an AI model. Labelling can be made with your own tools or not, online or offline, it can be segregation or segmentation.