Sara Bernardini: 'Designing Safe, Risk-Aware Autonomous Systems' | TAIS 2026

In this talk on autonomous systems in extreme environments, Sara Bernardini (Professor of Computer Science, University of Oxford) discusses how to design autonomous systems that are not only capable and intelligent, but also safe and risk-aware. Drawing on examples from underground mines, confined spaces, offshore energy infrastructure, and subsea domains (settings that are high-stakes, unstructured, data-poor, and difficult for humans to access directly) Sara argues that safety should be treated as a layered property of the autonomous architecture. Through examples from robotic exploration and offshore inspection, she shows how model-based autonomy can help systems reason about uncertainty, manage risk, adapt during execution, and remain understandable to human partners. Sara is a world-leading expert in automated planning, which involves equipping artificial agents with the ability to synthesise plans to reach specific goals. She is a member of the Executive Council of the Association for the Advancement of Artificial Intelligence (AAAI), where she leads the Partnership Committee, the International Conference on Automated Planning and Scheduling (ICAPS), and the Scientific Organizations Coordinating Council (AISOCC). Sara is an associate editor of the Artificial Intelligence Journal and the Journal of Artificial Intelligence Research. She was the managing guest editor of the Artificial Intelligence Journal Special Issue on ‘Risk-Aware Autonomous Systems’, a testament to her commitment to AI safety. This is a recording from TAIS 2026, a technical AI safety conference hosted at the Examination Schools, Oxford on 14 May 2026. Organised by Noeon Research in collaboration with Oxford Martin AI Governance Initiative, sponsored by MATS and Apart Research, with IASEAI and Foresight Institute as information partners. Thanks also to our OAISI volunteers. TAIS 2026 proceedings: https://tais2026.cc/proceedings Follow us on LinkedIn:   / tais-conference   Follow us on X: https://x.com/tais_2026