The CDT focuses on five horizontal research areas, from mobile autonomy to human-robot collaboration (see figure below), with an ambition to integrate and validate fundamental research in agriculture and food domains.

All of these topics may be applied to a diverse range of real-world challenges, from soil preparation, seeding, monitoring, scouting, weeding, spraying, selective harvesting and on-site grading, through to food processing, manufacturing and supply chain optimisation.

Other research areas within the CDT include Robot Vision; Robot Learning; Robotic Mapping; Robot Task Planning; Robot Navigation; Swarm Robotics; Systems Integration; Agri-Robotics; Food Manufacturing; Machine Learning and Applications of AI in Agri-food; fundamental mathematical models of robotics and Agri-food processes.

Research areas addressed by AgriFoRwArdS

AgriFoRwArdS addresses fundamental challenges in Robotics and Autonomous Systems (RAS) technologies for both agriculture and food production. The project should be based upon a business need and focus on one of the following RAS component technology areas, including but not limited to:

• Mobile autonomy: Agri-Food robots need to move in challenging dynamic, often GPS-denied and semi-structured environments with high precision. Autonomous mobility entails the integration of technologies for mapping, self-localisation and understanding of challenging farm and factory environments, dynamic path planning, precise motor control and locomotion, including safe operation in the presence of human workers.
• Manipulation and soft robotics: Manipulators are needed for a range of tasks, replacing dexterous human labour, reducing costs and increasing quality. Handling of delicate, unstructured objects such as food products requires new approaches to compliant and flexible manipulation. Example PhD topics might include vision- and tactile-guided handling and grasping tasks, and advanced functional materials for soft sensing and actuation.
• Sensing and perception: Machine vision and other modalities are needed for analysis of food products and sensor-guided control of robotic systems. Objectives might include classification of crops and weeds; phenotyping; quality analysis of food products; yield prediction; state estimation and modelling of farm or factory environments; detection, identification and tracking of human workers; etc.
• Fleet management: The true potential of robotics in agriculture and food production will be realised when different types of robots and autonomous systems are brought together in a systemic approach. Holistic approaches to fleet management are required, which fully integrate component methods for goal allocation, joint motion planning, coordination and control, as well as research on their integration and scaling to applications in agri-food.
• Human-robot collaboration: Many robotic applications will augment rather than replace human workers. Research may be needed into collaborative robotic systems or ‘co-bots’ that can work alongside human workers, for example, robots for fruit transportation working alongside human pickers, and to improve the safety of human-robot interactions in food production environments.

For further details on the research areas, groups and facilities available to students at the different partner institutes in the CDT, see the University Consortium page.