Sustainable manufacturing of AI hardware

The aim of this project is to develop a new form of neuromorphic systems that merge photonic, electronic and ionic effects, bringing new prospects for in-memory computing and artificial visual memory applications. This will be achieved upon developing photoelectric memories fabricated with more sustainable processes and greener materials.

AI is entering our everyday lives at an enormously growing pace but at a huge environmental cost. The energy required to run the AI algorithms that are used, for example, for image generation consume large amounts of energy. It is necessary to introduce alternative approaches with lower environmental impact.

Neuromorphic engineering offers a solution to this problem. The development of electronic devices that can realistically emulate biological neural networks holds promise for significantly reducing the energetic footprint and lowering the CO2 emissions generated by current AI hardware.

First, you will develop single devices that emulate biological synapses in the human visual system, capable of detecting (in analogy to the retina in the eye) and memorising or even processing images (like the visual cortex in the brain). Then, you will design and implement a novel neuromorphic optoelectronic array that will perform certain neuromorphic functionalities, e.g., pattern recognition tasks. Finally, you will assess the sustainability of this approach by applying life-cycle assessment techniques to each step of the fabrication process.

This novel electronic technology can effectively emulate synaptic weights and may be programmable both via light and voltage. This provides additional flexibility for implementing both synaptic weight updates as well as homeostatic effects. Furthermore, the technology relies on low-temperature processes and can thus be integrated on flexible substrates, which paves the way to incorporation of AI functionalities to wearable devices.

The outcomes of your research can have various applications, such as Internet of Things (IoT) devices for visual data communication, human/environment detection/tracking, Augmented/Virtual Reality, and more.

Entry Requirements

You must have a UK 2:1 honours degree or its international equivalent.

The AI for Sustainability CDT is multidisciplinary, and we welcome applicants from diverse disciplines, including but not limited to:

• engineering
• 
  
• social science
• 
  
• economics
• 
  
• business
• 
  
• computer science
• 
  
• mathematics
• 
  
• electronics
• 
  
• physical science.

You should have an interest in multidisciplinary research, and an aptitude for data analytics as well as other skills relevant to one or more of the core themes within the CDT. 

How to Apply

Apply online: Search for a Postgraduate Programme of Study (soton.ac.uk) 

• choose programme type (Research), 2025/26, Faculty of Engineering and Physical Sciences
• 
  
• choose 'iPhD AI for Sustainability' (Full time or Part time)
• 
  
• insert name of the supervisor Dr Dimitra Georgiadou, and the project title.

Applications should include: 

• your CV (resumé)
• 
  
• 2 reference letters
• 
  
• degree transcripts/certificates to date