About this webinar
Autonomous agents in defence are advanced AI systems that can perceive their environment, reason and independently execute multi-step tasks to achieve specific military objectives.
In this webinar, you'll hear about human-centred evaluation approaches for autonomous agents in safety-critical systems, before learning how this research has developed into a broader PhD programme on AI-enabled human-autonomy teaming in defence-relevant maritime contexts.
Our presenter, Benjamin Bowers, will summarise a systematic review of human-autonomy teaming research, highlighting how current evaluation practice remains fragmented, often relying on narrow or technology-centred measures rather than integrated evidence about team effectiveness indicators.
He'll then outline his PhD work translating these ideas into practical evaluation routes, including the development of a Human-AI Behaviour & Impact Toolkit (HABIT) measurement framework and questionnaire. He talk about workshop-based refinement with human factors researchers and practitioners, and experimental work on behavioural reliance calibration in AI-supported decision-making.
Benjamin will then discuss implications for early-phase Human-Systems Integration, defence testing and evaluation, and the practical challenge of deciding when an AI agent is genuinely supporting, rather than disrupting, safe and effective teamwork.
About the presenters
Samia Abdi (our chair) is a human factors specialist at BAE Systems.
Benjamin Bowers is a PhD Researcher in Human Factors Engineering within the Human Factors Research Group at the University of Nottingham. His research focuses on the human-centred evaluation of AI-enabled autonomous agents and human-autonomy teams in safety-critical systems. His work examined how autonomous agents are currently evaluated in the human-autonomy teaming literature and proposed routes toward more structured, human-centred assessment of team effectiveness. His ongoing PhD research builds on this work through the development of frameworks and guidance for evaluation and experimental studies of key human-autonomy teaming factors and human-centred AI design principles. Benjamin’s work aims to support more practical and theoretically grounded approaches to evaluating AI systems before they are integrated into complex operational environments.
Benjamin won the Best Paper Award when he presented this work as a poster at our Ergonomics & Human Factors 2026 conference in April.
