Marie Skłodowska-Curie PhD position on AI and robotics for sustainable agriculture

Back to jobs listing

INRAE presentation

The French National Research Institute for Agriculture, Food, and Environment (INRAE) is a major player in research and innovation. It is a community of 12,000 people with 272 research, experimental research, and support units located in 18 regional centres throughout France. Internationally, INRAE is among the top research organisations in the agricultural and food sciences, plant and animal sciences, as well as in ecology and environmental science. It is the world’s leading research organisation specialising in agriculture, food and the environment. INRAE’s goal is to be a key player in the transitions necessary to address major global challenges. Faced with a growing world population, climate change, resource scarcity, and declining biodiversity, the Institute has a major role to play in building solutions and supporting the necessary acceleration of agricultural, food and environmental transitions.

Work environment, missions and activities

You will be welcomed in the TSCF (Technologies et Systèmes d'Information pour les Agrosystèmes) unit. The TSCF unit at INRAE is dedicated to developing new technologies for agriculture, with a particular focus on agricultural and environmental robotics. Located in Clermont-Ferrand, in central France, this unit consists of 35 permanent researchers and is equipped with facilities for developing mobile robots, especially for off-road applications. The unit has a fleet of eight robots, equipped with sensors, designed for

various purposes and ranging in size from 10 kg to 6 tons. These robots feature either skid steering or Ackermann steering configurations. The laboratory also has 80 hectares of experimental fields, along with facilities for assessing robot efficiency. For the past 15 years, the TSCF team has been actively working on robotics for agriculture, developing numerous algorithms to control robots in dynamic and changing environments. Their work includes applications such as trajectory following, pedestrian tracking, and edge following.

Research project description

Agricultural robotics is central to the shift toward precision farming, which focuses on sustainable, efficient practices. These robots can reduce the environmental impact of farming, enhancing both productivity and conservation. However, they face significant challenges when deployed in dynamic, unstructured environments, including variable terrain, changing weather, and the need for real-time adaptability. As agricultural robots must operate in diverse and complex field conditions, it is crucial to integrate data from both local and landscape-scale sources to guide navigation and task execution.

The key challenge is improving robot’s ability to navigate and adapt to varied terrains, considering factors such as soil moisture, traction conditions, and topography. This proposal focuses on the fusion of local data—acquired from on-board sensors—and landscape-scale environmental data from UAVs, satellites, and weather stations. By combining these data sources, robots can develop a comprehensive understanding of their environment, which enhances task accuracy and decision-making in autonomous agricultural operations.

The research will develop a framework for integrating multi-modal sensor data from various sources. Drones and satellites provide aerial perspectives, while ground-based sensors capture localized environmental conditions such as soil moisture and temperature. These data will be combined into semantic maps that classify different areas of the field, allowing robots to make informed decisions about task planning and adaptation. The robot will use this information to navigate, managing obstacles and assess the suitability of the environment for specific tasks, ensuring safe and efficient execution.

In addition, the thesis will address the integration of landscape-scale data, such as weather forecasts and topographic maps, to complement the local data from the robot. The objective is to allow robots adapting in real time to environmental changes, adjusting navigation parameters or control algorithms accordingly. For example, a robot might alter its path or delay an operation if soil moisture is too high, preventing potential damage to crops or equipment. A core challenge of this research is the development of machine learning models that can effectively process and integrate data from these diverse sensor systems. Deep Learning techniques will be explored to handle the spatial, spectral and temporal dimensions of the data.

Finally, the effectiveness of this multi-sensor framework will be validated through experimental field trials. These tests will evaluate how well robots can navigate, plan tasks, and adapt to environmental changes in real-world agricultural settings and tasks to be achieved. By integrating both local and landscape-scale data, the robots will demonstrate enhanced autonomy and decision-making capabilities, facilitating the transition to a more sustainable and data-driven farming model.

Expected results

• Identify and evaluate the types of data required for safe and efficient planning of agricultural robotics tasks.
• Design and implement a framework for data fusion, using Deep Learning techniques to combine various data modalities, including onboard sensors and landscape-level information.
• Build dynamic mapping systems capable of adapting in real-time to environmental changes, providing the robot with up-to-date terrain and environmental information.
• Create an ontology to define environmental relationships.
• Validate the proposed frameworks and algorithms through real-world experimental trials in diverse agricultural settings.
• Publish results in international conferences and high-impact journals, contributing to the field of agricultural robotics and multi-sensor data fusion.

Upload your application (single PDF smaller than 10 MB) before the deadline at: https://forms.gle/8dDE9p4WnRzk5EXMA.

Your single PDF file should contain:

• A detailed CV (including your publications and/or projects with links to online material like public repositories, articles, etc.).
• A motivation letter explaining your interest in the project and your relevant skills (1 page).
• BSc and MSc academic transcripts.
• Degree certificates
• Links to the PDF of any other relevant documents (Master's thesis, portfolio, etc.).
• Name, email, and phone number of at least one referee (professor or researcher you have worked with) to whom we can reach out; no letter required.

Work with robots possibly in outdoor environments.

Training and skills

Recommended training: You should have a MEng/MSc degree, or equivalent, in computer science, robotics, mathematics, physics, or related fields.

Knowledge required: Strong programming skills (preferably in Python or C++).

Appreciated experience: Some practical experience on robotics and machine learning.

Skills sought: Motivation, sense of responsibility, autonomy and problem-solving skills are highly desirable.

Specific requirements: Your application should respect the AIGreenBots general requirements and eligibility criteria. These include that the candidate must not have resided or carried out his/her main activity (work, studies, etc.) in France for more than twelve months in the three years immediately prior to the call deadline.

INRAE's life quality

By joining our teams, you benefit from (depending on the type of contract and its duration):

- up to 30 days of annual leave + 15 days "Reduction of Working Time" (for a full time);
- parenting support: CESU childcare, leisure services;
- skills development systems: training, career advise;
- social support: advice and listening, social assistance and loans;
- holiday and leisure services: holiday vouchers, accommodation at preferential rates;
- sports and cultural activities;
- collective catering.

The Cézeaux site is served by tramway line A, and is also equipped with parking facilities and services dedicated to cycling.

How to apply

I send my CV and my motivation letter