We are engaged in research focused on developing tools and methods to enhance the efficiency and sustainability of agricultural operations through the use of autonomous agricultural machinery. Our efforts are structured around four main areas:

Robotic Mission & Planning

We work on robotic mission planning to optimize agricultural tasks. Specifically, we develop strategies for planning crop-specific routes and tasks to optimize machine paths in the field, taking into account crop characteristics and environmental constraints. We also develop algorithms for trajectory planning and tool operation to optimize machine movement and ensure precise tool deployment. Additionally, our research covers planning and coordination of fleets of mobile platforms, working on the synchronized management of multiple autonomous machines to efficiently perform agricultural tasks within plots.

Robust Localization

We develop advanced localization systems to enable autonomous machines to interact effectively with their environment in diverse and often complex contexts. To this end, we work on multi-modal, robust, and season-invariant localization solutions that can adapt to changing environments throughout the year, regardless of weather or visibility conditions. To ensure maximum flexibility, we also design indoor–outdoor localization systems that allow smooth transitions between indoor environments, such as farm buildings or greenhouses, and outdoor environments, such as fields and pastures, while maintaining optimal precision and robustness. In addition, we actively contribute to advancements in simultaneous localization and mapping (SLAM), by developing techniques specifically adapted to agricultural settings, which are often characterized by dense vegetation or moving obstacles. Finally, to address the challenges of dynamic environments, we explore innovative approaches to decentralized and shared localization, where multiple autonomous machines collaborate by fusing their data to create coherent maps and collectively adapt to changes in their surroundings.

Control of Mobile Platforms & Tools

We work on mastering and adapting autonomous machines to the diverse conditions found in agricultural environments. As part of this effort, we develop sensor-referenced control systems (using GPS, LiDAR, and vision) to ensure precise control of machines and their tools, enabling the execution of complex agricultural tasks such as hoeing and weeding while preserving crop integrity. We also focus on terrain adaptation and traction control, by designing systems that allow machines to adjust to soil conditions, including moisture, roughness, and slope, ensuring accurate task execution despite variable ground conditions. Another important area of our research involves the coordination of ground and aerial mobile platforms, aiming to synchronize the movements of terrestrial and aerial robots to enable cooperative agricultural operations. We also develop solutions for handling deformable objects, making it possible to efficiently manipulate elements such as plants and flexible agricultural materials using robotic arms mounted on agricultural vehicles. Finally, we design control architectures for managing redundancy and tool control, integrating additional degrees of freedom; whether internal to the robots, such as active suspensions, or external, such as robotic arms and actuated tools.

Robotic Integrity

Ensuring the integrity of autonomous mobile platforms is a major challenge, particularly in the varied and sometimes difficult environments encountered in agriculture. To meet this challenge, we actively contribute to the functional safety of these systems by helping to design localization and control solutions that incorporate regulatory and normative constraints, as well as their future developments. This approach not only ensures the integrity of autonomous operations but also reduces the risks associated with the deployment of autonomous agricultural machinery. In addition, to prevent incidents related to terrain conditions, we have developed anti-rollover and traction management systems that detect hazardous ground and automatically adjust control parameters. These systems help prevent rollovers and loss of grip, thereby improving machine safety.

Finally, our work on traversability and obstacle avoidance focuses on terrain analysis and the identification of passable or high-risk areas. Through environmental perception, this research aims to avoid obstacles and enhance the navigation capabilities of autonomous machines in diverse contexts.

Through these research areas, we aim to address the growing challenges of modern agriculture by integrating innovative solutions that enhance performance, safety, and natural resource management. Our commitment to these topics contributes to shaping the future of sustainable and intelligent agriculture.