Bio-inspired Systems, Neuromorphic Hardware, Unconventional Computing

Applicants are invited in one, or a combination of, the following areas:

Bio-inspired Systems

Wonders in Nature, including our own brains, have fascinated scientists and en- gineers for decades if not centuries. There has been huge international research interest in studying all kinds of nature-inspired computation and its applications, including evolutionary computation, neural computation, and swarm intelligence. The vast majority of such research is carried out through software implementation and simulation.

Evolvable Hardware is a unique research field that poses research challenges in theories, algorithms, software, and hardware. It is not just about how to implement smart algorithms in hardware, but more about how to make hardware itself smarter by adapting its structure and functionality online in a dynamic and uncertain environment. Such a vision of making hardware soft” requires the hardware substrate to be sufficiently flexible so that its structure can be changed online, i.e., at run-time. This is a huge challenge because silicon-based hardware is not easily changeable. This is also an opportunity because it encourages us to think about and explore other hardware substrates that could perform computation in a more flexible and adaptable way, and investigating novel approaches in this space is the aim of this research project.

Neuromorphic Hardware

Spiking Neural Network (SNN) building blocks of synapses and neurons have been demonstrated in the individual small scale, and sophisticated bio-inspired models have been developed for fault-tolerance and system management. Now is the time to combine and push the boundaries of these technologies to build the next generation of electronic systems that provide ultimate reliability, that can sense, adapt, classify, and even predict anomalies.

This project aims is to develop this scalable embodied Nervous System, including new design methodology and hardware architecture, that can be exploited to design reliable electronic systems by introducing “self-aware” subsystems comprised of networks of artificial spiking neurons, capable of sensing their state of operation, coupled with adaptive capabilities in recognising anomalies and triggering appropriate alerts and repairs.

Unconvontional Computing

Classical digital computing is power hungry, fragile, and hard to interface to the real world. Reservoir computing (RC) can help overcome these issues, particularly by being able to perform embodied computation that can directly exploit the natural dynamics of the substrate, thereby dramatically reducing power requirements and providing a natural fit to certain computational tasks. Introduced more than a decade ago, RC has proven to be an efficient approach for signal processing and dynamical pattern recognition, and state-of-the-art performance has been demonstrated in both simulation and physical implementation. These two research projects will help take RC to the next level: designing, building, applying and analysing complete, efficient and feasible RC architectures: a networked “reservoir of reservoirs” (RoR) platform.

This project will investigate unconventional computing substrates (e.g., physical systems, nano materials, analogue electronics) to support efficient, flexible and feasible multi-reservoir computers. The aim is to demonstrate novel approaches on a range of applications, starting from a smart microphone implementing noise cancelling, sound source isolation and speaker identification, and building up to a universal “edge processor” that can process data from diverse sensors (different timescales and modalities), trainable for different tasks and contexts.

Based at York.

How to Apply:

Applicants should apply via the University’s online application system at https://www.york.ac.uk/study/postgraduate-research/apply/. Please read the application guidance first so that you understand the various steps in the application process.