Self-repairing Hardware Paradigms based on Astrocyte-neuron Models

Lead Research Organisation: University of York
Department Name: Electronics


The human brain is remarkable in its ability to self-repair, for example following stroke or injury. Such self-repair results from a range of distributed and fine-grained mechanisms which act in tandem to ensure that the neurones (the basic building blocks in the brain) continue to function in as close to a normal state as possible.

In contrast modern electronic systems design typically relies on a single controller or processor, which has very limited self-repair capabilities. There is a pressing need to progress beyond current approaches and look for inspiration from biology to inform electronic systems design.

Recent studies have highlighted that interactions between astrocytes (a type of glial cell) and neurones in the brain provide a distributed cellular level repair capability where faults that impede or stop neuronal firing can be repaired by a re-adjustment of the local weights of connections between neurones in the brain.

This project aims to exploit these recent findings and develop a new generation of self-repairing algorithms by taking inspiration from these results to design a new generation of "astro-centric" algorithms. To achieve this we will include components representing both neurones and astrocytes in our electronic systems and model the interactions between these in such a way as to capture the distributed repair capabilities seen in the biological system.

Planned Impact

Our project is "blue-skies research" at the cutting edge of bio-inspired design for next generation electronic systems. The project aims to develop computational tools, algorithms and robotic demonstrators to implement distributed, fine-grained repair in spiking neural networks that take inspiration from recent findings regarding biological brain repair mechanisms.

Economic impact.
We envisage that our project will have longer term economic impact. Uptake of our ideas by the UK electronics design industry has the potential to increase economic activity through additional exploitation routes for novel design principles. The UK is well placed to benefit from this research as 40% of European design houses are based in the UK. Exploitation routes are discussed in the pathways to impact document.

Societal impact
The research targets application areas in robotics as demonstrators. This choice was motivated in part to increase potential impact beyond the scientific community. Demonstration of robotic systems where a brain inspired electronic control system is able to respond to a range of fault scenarios has the potential to generate considerable publicity and associated impact to highlight bio-inspired Engineering research in the UK. Additionally, such reliable robots will be of benefit in the domain of assisted living. The research at Ulster will assist in the understanding of brain function/dysfunction and therefore impact in general mental health and well-being.
Description Recent results from Neuroscience have highlighted the role of astrocyte-neuron interactions in regulating neuronal activity in the brain. The aim of this project was to take inspiration from these findings to develop electronic systems with embedded fine-grained self repair capabilities based on spiking astrocyte-neuron networks.

Key findings
1. We have developed computer models of astrocyte-neuron interactions which capture the self-repair capabilities that result from astrocyte-neuron interactions.
2. We have mapped these complex software models to a range of representations suitable for use in electronic hardware.
3. We have implemented these representations in Field Programmable Gate Array systems and verified the fault tolerant behaviour in hardware.
4. We have pioneered the homeostatic spiking neuron model as a basic building block in fault tolerant spiking neural networks.
5. We have developed robot lab infrastructure to support use of single and multiple mobile robots to demonstrate principles of fault-tolerant hardware.
6. We have demonstrated hardware fault tolerance in mobile robots during a follow-the-leader task and during an obstacle avoidance task - the robot continues to function when faults are induced in the hardware control system.
Exploitation Route These findings could be taken forward during a follow on research project to further develop the approach of fault-tolerance using spiking astrocyte-neuron networks. Opportunities for commercial applications include proof-of-principle demonstrators, for example, fault-tolerant sensor systems as a means of advancing the Technology Readiness Level of spiking astrocyte-neuron networks.
Sectors Aerospace, Defence and Marine,Digital/Communication/Information Technologies (including Software),Electronics

Description The natural world has examples of living organisms which exhibit desirable behaviours from the perspective of electronic systems. One of these behaviours is fault tolerance. The SPANNER project pioneered a novel approach to distributed fine-grained fault tolerance inspired by findings on astrocyte regulation of neural activity in the brain. The ideas of self-repair in distributed computing systems has potential to impact on current practice and develop new approaches to the design of electronic systems. Our approach has recently been presented to a wide audience through the Physics of Life network, leading to discussions on development of the ideas from a broader community of scientists outside Engineering. The work was presented as an invited contribution to techUK's Future of Compute Week 2022 in an event on neuromorphic computing to explain why the UK tech sector needs to understand neuromorphic computing, and the benefits it may one day unlock The work is also under discussion with a number of leading companies in the areas of autonomous systems in mission critical scenarios. These developments will be described in future updates.
First Year Of Impact 2020
Sector Aerospace, Defence and Marine,Healthcare
Impact Types Societal

Description Neuromorph-what? What is neuromorphic computing and what does this mean for the future of compute?
Geographic Reach Multiple continents/international 
Policy Influence Type Influenced training of practitioners or researchers
Description Autonomy in hazardous scene assessment
Amount £78,500 (GBP)
Funding ID ACC 101169 
Organisation Defence Science & Technology Laboratory (DSTL) 
Sector Public
Country United Kingdom
Start 01/2016 
End 06/2016
Description Life Science Centre, Feb 2016 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Public/other audiences
Results and Impact Contribution to Life Science Centre event to promote science and technology, held at Life Science Centre Newcastle. Event used miniature swarm robotic system developed at University of York as part of the associated EPSRC project.
Year(s) Of Engagement Activity 2016
Description Physics of Life - Physics of Brain workshop, invited talk: Neuromorphic computing systems: spiking neural networks, astrocyte-neuron networks 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Professional Practitioners
Results and Impact There is increasing and intense interest in the physical science, and mathematical modelling of aspects of brain and mind that animate most multicellular species. Electrophysiology of neurons and nerves, on the one hand, and modelling and application of neural nets, on the other, are examples of long-standing research programmes at or motivated by the molecular level of structure. Similarly the field of psychophysics has developed sophisticated non-invasive methods for investigation of perception, memory and learning in humans. Yet there is also huge scope for the application of methodologies both experimental and theoretical, from branches of science not currently associated strongly with neuroscience. Advanced imaging, electromagnetic methods, adaptive optics are experimental examples, statistical mechanics and stochastic non-linear systems are promising theoretical tools. The application of novel methods from the physical sciences to brains was identified strategically by the EPSRC Life Science Interface, and has remained a strategic imperative, included in the Roadmap for Physics of Life produced by PoLNet under the key theme of 'information flow in biological systems'. Yet this workshop represents the first PoLNET workshop to focus on neuroscience. The workshop is also the first of a series of interdisciplinary workshops, supported by the Rosetrees Trust and embedded within PoLNET, which focus on key clinical challenges (see Physics of Medicine). The overall question is, what current challenges in brain science could be met by bringing new physical science methods, experimental and theoretical, to bear, together with biomedical, psychological and biological science.
Year(s) Of Engagement Activity 2020
Description Raspberry Pi Challenge 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach Local
Primary Audience Undergraduate students
Results and Impact Each year new undergraduate Computer Science students are sent a free Raspberry Pi after their places are confirmed. Students are given access to York's VLE (Visual Learning Environment) where they are able to sign up for either, or both, of the two challenges. The challenges culminate in an event at the end of the first week at University.
Year(s) Of Engagement Activity 2017
Description SPANNER Workshop 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Professional Practitioners
Results and Impact This was an international workshop with speakers from Europe and USA to discuss and reflect on the findings from this project. There were a number of invited keynote talks and informative discussions in the general area of bio-inspired design.
Year(s) Of Engagement Activity 2018
Description UKESF/Headstart Summer School, University of York, July 2016 
Form Of Engagement Activity Participation in an open day or visit at my research institution
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Schools
Results and Impact The UKESF-sponsored Summer School for 2016 took place 17-21 July at the University of York, and was organised in partnership with Headstart EDT and York's Department of Electronics. 80 STEM-focused Year 12/S5 students attended the residential week to get a first taste of university life and learn more about studying Electronics at degree level, with a variety of lectures, labs, visits and networking opportunities. Sessions ranged from 'MEMS Biosensors and the Potential for Improving Healthcare' to 'From the Ancient Greeks to the iPhone: a Brief History of Communications', as well as hands-on challenges that dealt with design and robotics. One of the activities was a robot lab challenge using the swarm robotic system developed as part of the associated EPSRC project.
Year(s) Of Engagement Activity 2016