Research

This domain focuses on using robotic and autonomous systems to protect man-made assets and public spaces and to support safety and security operations. Typical assets include bridges, tunnels, power plants, pipelines, transport networks, ports and border installations. Robots and drones are used for structural inspection and maintenance, perimeter and border surveillance, support to law-enforcement and first responders, and operations in hazardous built environments (e.g. CBRN, industrial accidents). In our research, we are especially interested in how to deploy teams of unmanned vehicles and sensor systems to patrol extended coastal and land borders, monitor critical sites, detect hazardous releases or intrusions early, and provide operators with timely, fused situational awareness. Key challenges include operating safely near people, complying with strict safety and security regulations, integrating with command-and-control and emergency-management systems, and maintaining resilience against adversarial or malicious behavior.

Precision agriculture uses sensing, robotics and data analytics to optimize crop and livestock management at fine spatial and temporal scales. Research topics include UAV and ground-robot sensing for crop-health monitoring, weed and pest detection, yield estimation, variable-rate application of inputs, and decision-support systems that fuse multi-modal data. Our work typically focuses on how unmanned aerial vehicles can efficiently scan large fields, how to translate aerial images into maps of weeds, stress or yield, and how to turn these maps into actionable, easy-to-use recommendations for growers. The overall goal is to increase productivity and sustainability while reducing inputs such as water, fertilizer and pesticides, with challenges related to highly variable outdoor conditions, robust perception for vegetation, and tools that are affordable and usable for farmers of different scales.

This domain covers autonomous vehicles and intelligent transport systems, including cooperative, connected and automated mobility (CCAM) and logistics applications. Topics include perception and localization for road vehicles, motion and trajectory planning in mixed traffic, vehicle-vehicle and vehicle-infrastructure communication, fleet and logistics optimization, and safety and validation frameworks. In our research, we often tackle problems such as planning safe motions for automated vehicles around dynamic obstacles, coordinating mixed fleets of human-driven and autonomous vehicles, and scheduling electric-vehicle charging and delivery operations so that mobility demand is met with minimal congestion and energy cost. Key challenges include operation in dense, unpredictable urban environments, large-scale testing and validation, cybersecurity, and alignment with regulations and user acceptance.

This domain applies robotics and multi-robot systems to observe, understand and protect the natural environment, with a strong emphasis on natural disasters and pollution. Robots and autonomous vehicles (UAVs, USVs, AUVs, UGVs) are used to monitor ecosystems and biodiversity, track air, water and soil quality, and provide early warning and damage assessment for wildfires, floods, storms, landslides and other environmental hazards. A key focus is on detecting, mapping and characterizing pollution sources, such as illegal landfills, landfill emissions, industrial spills and urban air-pollution hotspots. Our contributions often concentrate on mission planning and data analysis for such systems, for example when using aerial and underwater robots to construct detailed maps of seafloors, large landfills or forest areas, estimate volumes and types of waste or biomass, and support early detection and assessment of environmental crimes or fire risk. Research challenges include planning missions over very large or remote areas, ensuring long-term autonomous operation in harsh conditions, achieving high spatial and temporal resolution in measurements, and turning raw environmental data into actionable information for conservation, climate adaptation and environmental law-enforcement.

Cultural Heritage Preservation & Digitization uses robotics, advanced sensing and AI to document and protect artifacts, monuments and collections, while making them more accessible and engaging for experts and the public. The central problem is to capture and maintain high-fidelity digital representations of fragile or unique objects—such as textiles, sculptures or architectural details—and the spaces that house them, without disturbing or damaging the originals, and then to turn these digital assets into tools for conservation, study and visitor experience. In our work this typically involves mobile platforms and robotic scanning setups that can manoeuver around exhibits, collect multi-view and multi-modal data under realistic museum constraints, and support the creation of detailed 3D or 4D “digital twins” that curators, conservators and educators can use for monitoring condition, planning restoration or building interactive experiences on-site and online. This domain faces challenges such as operating in tight or crowded spaces, dealing with complex materials and lighting, managing and curating very large datasets, and ensuring that the resulting digital heritage is usable, inclusive and sustainable for cultural institutions.

Human-Robot Collaboration & Assistive Robotics studies how robots can safely share physical and social spaces with people, supporting them rather than replacing them in work, home and care settings. The focus is on robots that work side-by-side with humans in tasks such as handling objects, guiding or accompanying people through buildings, and assisting older adults or people with reduced mobility in their daily routines, often as part of a sensor-rich ambient environment. Typical problems include understanding human intent and state, coordinating motion and task execution with human partners, and providing intuitive ways for people to give commands and feedback while maintaining safety, privacy and comfort. In our lab this usually translates to mobile and socially interactive robots that must move in cluttered real environments, adapt to different users' abilities and habits, and remain trustworthy and easy to supervise over long periods of time, which raises challenges around robust perception, real-time interaction, acceptance and ethical deployment.