PhD in Computer Science

Award Summary 100% fees covered, and a minimum tax-free annual living allowance of £20,780 (2025/26 UKRI rate). Additional project costs will also be provided. Overview How can we revolutionise intelligent systems that still rely on fixed rules and structured commands? For example, if asked to “pick up the red book on the shelf,” a traditional robot would need the exact location and object details programmed in advance. If anything changes, it may fail. This project explores how to build more adaptable systems using vision-language-action (VLA) models. These combine computer vision (to see), natural language understanding (to interpret instructions), and action generation (to respond), enabling robots to perceive, reason and act flexibly. The models will be trained on simulated datasets to learn general behaviours, then fine-tuned for specific tasks. The aim is to create intelligent agents that could demonstrate human-like understanding and adaptability if applied to real-world scenarios.