PhD Studentship: Evolving World Models for Self-Adaptive and Autonomous Systems using Evolutionary Reinforcement Learning University of Exeter in United Kingdom

[£20,780 per year stipend, payment of Home tuition fees, and a Research Training Support Grant of £5,000 over 3.5 years.]

This fully funded PhD studentship at the University of Exeter offers an exciting opportunity to advance the field of self-adaptive and autonomous systems through the development of evolving world models using evolutionary reinforcement learning techniques. The project addresses the challenge of maintaining accurate models in dynamic and uncertain environments, where traditional expert-specified models often fail due to rapidly changing conditions. Applications span crowd management, large-scale communication networks, healthcare, and environmental monitoring, with real-world scenarios such as mass gatherings and emergency events highlighting the need for robust, real-time adaptation.

The research will focus on combining evolutionary algorithms and reinforcement learning (RL) to continuously learn, adapt, and refine transition and observation models. Key objectives include investigating RL for constructing and updating probabilistic world models, developing hybrid frameworks that integrate evolutionary search with RL’s incremental updates, and ensuring robustness to various sources of uncertainty. The project aims to deliver novel methods for refining expert-specified models based on environmental feedback, improved decision-making strategies for adaptation in uncertain conditions, and experimental validation in real-world inspired scenarios.

As a doctoral researcher, you will gain hands-on experience in implementing RL and evolutionary optimisation, with a use case in dynamic crowd and network management provided by the industrial partner INOCESS Inc., France. The partnership offers co-supervision, access to relevant data, opportunities for field testing, and potential hosting at INOCESS facilities. This collaboration ensures that your research will have practical impact and relevance to industry needs.

Funding for this studentship includes a stipend of £20,780 per year, payment of Home tuition fees, and a Research Training Support Grant of £5,000 over 3.5 years. Applicants should have a strong academic background in Computer Science, Mathematics, Engineering, or related disciplines, with experience or interest in AI, RL, evolutionary algorithms, or autonomous systems. International candidates may need to provide proof of English language proficiency.

The application deadline is 12 January 2027. To apply, visit the University of Exeter funding award portal and ensure you meet the entry requirements. For further information or project-specific enquiries, contact Dr Huma Samin at [email protected].

This studentship provides a unique opportunity to contribute to the advancement of adaptive and autonomous systems, equipping you with cutting-edge skills and experience in a rapidly evolving research area.