Bath Engineering & Design Studentships: Causal Artificial Intelligence to Empower Resilient Low-Carbon Energy Systems

This project is one of a number in competition for a studentship from the Faculty of Engineering and Design at the University of Bath. If successful, this studentship is expected to commence 04 October 2027.

An alternative start date may be possible if agreed with your intended supervisors and the Doctoral College.

Achieving net zero requires a fundamental transformation of the electricity systems—from centralised, fossil-fuel-based infrastructure to a dynamic, renewable-rich network. This transition is stretching legacy assets beyond their design limits while exposing the system to new vulnerabilities: climate extremes, ageing infrastructure, cyberattacks, and the integration of distributed energy resources, electric vehicles, and storage. In such a complex, tightly coupled system, even small disturbances can escalate into widespread failures. The consequences are already clear: the 2019 UK blackout, triggered by two unrelated failures; in 2025, a transformer fault at Heathrow caused fires and widespread outages; and the 2025 Iberian Peninsula blackout, rooted in solar generation voltage instability, escalated into a full system failure. These outage events highlight three urgent questions in energy system resilience:
• What were the true root causes of these outages?
• What hidden vulnerabilities do allow local faults to cascade?
• Could earlier, smarter interventions have prevented or contained the outages?
Answering these questions is not possible with existing models and tools. Energy systems are nonlinear, dynamic, and deeply interdependent—not just technically, but also with environmental, geographical, and socioeconomic factors. What is urgently needed is a new paradigm: a causal, knowledge-driven AI framework that can integrate heterogeneous data, fuse domain knowledge across disciplines, uncover root causes, and run “what-if” scenarios. This PhD project will address these challenges by developing a Causal AI framework underpinned by a knowledge graph to transform how energy system resilience is understood and safeguarded.
• Causal discovery at scale: Build a unified knowledge graph integrating operational, environmental, and cyber-physical data, enabling causal inference to reveal hidden pathways of risk and vulnerability.
• Causal digital twin: Embed physical laws and engineering constraints into structural causal models to simulate cascading failures and evaluate counterfactual “what-if” scenarios.
• Autonomous resilience: Train AI agents on the causal digital twin to propose adaptive strategies that strengthen resilience against climate extremes and emerging risks.
By combining knowledge graphs, causal reasoning, and AI, this project will provide interpretable, scalable, and actionable tools to secure the reliability of future low-carbon energy systems and accelerate the net-zero transition.