Professor Tim Norman
CDT Director
The MINDS Centre for Doctoral Training trains researchers to solve one of the defining problems in modern AI: how do you build intelligence into the physical hardware that the world runs on? Energy consumption, device constraints, processing speed; these are not obstacles to route around. They are the research.
MINDS brings together expertise in machine intelligence and nanoelectronic devices and systems, combining AI algorithm development with the materials science, device physics and fabrication engineering needed to make that intelligence work at scale in sensors, semiconductor systems, connected devices and beyond. Based at the University of Southampton's School of Electronics and Computer Science and Zepler Institute for Photonics and Nanoelectronics, MINDS researchers work at the intersection of two fields that are rarely trained together.
We bring together world leading AI and electronics expertise in this unique centre to develop and support a new generation of scientists innovating future energy-efficient, hardware-enabled AI to the benefit of our society and economy.
Our research
MINDS research is organised around four interconnected themes.
Each addresses a different dimension of the same underlying challenge: making AI work effectively, efficiently and reliably in the real world not just in theory, and not just in the cloud.
Agent-Based Adaptive Systems
How do groups of autonomous AI agents communicate, coordinate and make decisions; especially when they cannot rely on centralised control? Research in this theme explores the languages agents develop, the protocols they use, and how those behaviours can be made more efficient, interpretable and robust.
Embedded AI
How do you run sophisticated AI on hardware that is small, low-power, or physically constrained? This theme focuses on the design and implementation of AI systems that work within the real limits of embedded and edge devices from sensors to integrated circuits.
Task-Optimised Devices and Systems
How do you design hardware and systems specifically for AI tasks, rather than adapting general-purpose computing? Research here spans the development of device architectures, fabrication techniques and system-level approaches that are optimised from the ground up for machine intelligence.
Nanoelectronic Technologies for AI
AI processing makes fundamentally different demands on hardware than conventional computing favouring architectures built around multiply-accumulate operations, adjustable memory and co-located processing and storage. Research in this theme investigates alternatives at the materials and device level, including emerging technologies that exploit nanomolecular effects to store and process information more efficiently.
How we work
MINDS uses a cohort model.
Researchers join as a group, move through a shared first year of training together, and then develop individual research projects with dedicated supervisory teams. That structure is deliberate. Working across disciplines and alongside researchers tackling different problems within the same broad field produces a kind of thinking that subject-specialist programmes rarely develop.
Researchers are admitted to the MINDS programme rather than to work with a specific supervisor from the start. That gives the first year its purpose: time to explore the literature, develop ideas, and find the supervisory combination usually including an industry advisor that will support the best possible research. In hindsight, this approach consistently produces more confident, more independent researchers.
The CDT's connection to industry is structural, not occasional. Partners contribute to research direction, co-supervise projects, host placements, and engage with outputs throughout the programme. That means MINDS researchers encounter real data, real constraints and real decision-making from early in their training not as a finishing step, but as part of how the research is shaped.
MINDS is also part of a wider network. The University of Southampton is a member of the Alan Turing Institute, and MINDS researchers have access to that national network alongside the CDT's own connections across the semiconductor, AI and electronics ecosystem.
Our team
Professor Harold Chong
CDT Deputy Director
Dr Bahar Rastegari
Training & Cohort Development; Agent-Based Adaptive Systems theme leader
Dr Christine Evers
Embedded AI Theme Leader
Professor Rob Maunder
Impact Acceleration; Task-Optimised Devices and Systems Theme Leader
Dr Dimitra Georgiadou
Advertising & Recruitment; Deputy Impact Champion
Dr Reena Pau
Outreach and Diversity co-ordinator
Dr Stuart Middleton
STARS programme coordinator
Emma Miles
CDT Programme Manager
