My research focuses on designing and engineering intelligent and autonomous systems for complex, real-world applications and lies at the intersection between different areas: artificial intelligence, advanced robotics and mathematical optimisation.
My goal is to create machines that can seamlessly support humans in undertaking sophisticated tasks that, going beyond perception and pattern matching, require deliberation at the cognitive level.
I have done extensive research on AI planning and scheduling, multi-agent systems, probabilistic reasoning, combinatorial optimisation and human/machine interaction.
My expertise is in cutting-edge technology for challenging environments. I have worked in several domains, including space mission operations, nuclear decommissioning, mining, underwater operations, surveillance, offshore energy and disaster response.
Currently, I am the co-PI of two Innovate UK projects concerning AI and robotics in extreme environments:
SoAR (Squad of Adaptive Robots) (£2m) aims at delivering a fundamental shift in the use of marine robotics for seabed inspection and exploration via the networking of heterogeneous fleets of AUVs and uncrewed surface vehicles overseen by a fleet-level autonomy engine capable of adaptive mission planning.
Demeter (£2m) uses energy-harvesting, uncrewed, intelligent vehicles to provide a persistent and autonomous subsea data retrieval and analysis platform to enable more cost-effective predictive maintenance of subsea assets, such as renewable and non-renewable offshore resources.
In 2019-21, I was the co-PI of three Innovate UK projects concerning AI and robotics in extreme environments:
Connect-R, the second-largest project ever funded by IUK (£6m), concerns a self-building, modular robot that assembles itself to provide structure in unstructured hazardous environments.
MIMRee is a large project (£4.2m) that aims to develop the first fully autonomous multi-robot platform for the inspection, maintenance and repair of offshore wind farms.
Prometheus (£2m) focuses on building an autonomous, reconfigurable drone for the inspection of confined spaces, such as underground mines.
In all those projects, I have been responsible for the design and implementation of the robots' autonomy architecture, which makes them capable of undertaking sophisticated tasks without close human supervision.
Currently, I am also the co-PI of the Innovate UK project VersaTile (£2m), which aims at transforming warehousing and distribution operations by developing an AI-powered modular robotic system that can continuously increase volumes and throughput whilst minimising warehouse space and labour utilisation, and the project SMaILE (£1m), funded by Compagnia di San Paolo, which focuses on teaching AI to young adults. Finally, I lead a Leverhulme grant on using game theory for large-scale, hybrid decision-making problems.
Learn more about my Projects.
Theory and Practice
While seeking to build systems capable of working in the real world, I have maintained a keen interest in theoretical research. My goal has always been to elicit fundamental research questions from practical problems and, at the same time, to transfer theoretical insights into practical domains.
My work on sequence submodularity, Through the Lens of Sequence Submodularity, has recently received a Best Paper Honourable Mention Award at the top conference in automated planning and scheduling (ICAPS-20). It has also been published by the journal of Artificial Intelligence (AIJ - 9.088 impact factor).
Learn more about my Publications.
I am the Associate General Chair of the 37th AAAI Conference in Artificial Intelligence, Managing Guest Editor of the Artificial Intelligence Journal Special Issue in Risk-Aware Autonomous Systems: Theory and Practice, co-Chair of the Doctoral Consortium at ICAPS-22, and Publicity Chair of AAMAS-22.
Learn more about my Activities.