eFuturesXD - crossing the boundaries

Lead Research Organisation: Newcastle University
Department Name: Electrical, Electronic & Computer Eng

Abstract

An EPSRC workshop (ICT research - the next decade) held in October 2010 recognised the need for ICT communities to shape future capability. In order to help achieve this goal in the electronics community over the coming years, eFuturesXD will offer awards to support UK academics in electronics as they initiate collaborative research across their discipline boundary. This tends to be research of a high risk, high return kind. Modest awards allowing immediate progress will be an efficient way to sift numerous options and to promote the best ideas. The outcome will be a collection of short studies, which will include varying levels of promise. This will help inform the electronics community where ICT research should be heading. By offering a response to award requests within one month of applying, this cross-disciplinary account will give the agility necessary to compete on the global stage. Four kinds of award will be offered: travel award (< £5k); meeting award (< £10k); facility award (< £20k); and staff award (< £50k). These awards will offer the opportunity to scope new initiatives rapidly and to assemble the best team for longer larger research programmes. The cross -disciplinary account will be facilitated by the existing electronics community network eFutures, building on its established infrastructure and network of technical expertise. It will also be a way of adding value to the eFutures network. The cross-disciplinary account will be publicised through community meetings and the eFutures web pages. A condition of obtaining an award < £15k will be to give presentations at community events. By linking the cross-disciplinary account to eFutures it will be possible to help build the best consortia for longer larger research programmes using the network's links to the academic and commercial sectors. The application process will involve completing an on-line form which will cover summary, objectives, academic partners, funding requested and strategic alignment. Rapid assessments will be made by two experts: one from electronics and one from the other relevant discipline. A review panel will be held monthly when decisions will be made and then communicated to applicants. The eFutures steering group, comprising academic, commercial and public sector representatives, will review the distribution and balance of awards across strategically important areas where the UK may take a future lead. Success will be seen by measuring the follow-on funding by those obtaining awards. It will also be judged by looking at the overall ICT portfolio. Examples of success include a more coherent range of research topics within electronics and a visibly stronger engagement between electronics and the rest of ICT. To assess the effectiveness of awards a follow-up check will be made at least 6 months after each award is complete.

Planned Impact

The eFuturesXD account sets out to encourage and support the initial stages of cross-disciplinary research through a variety of small awards. The intended outcome will be a number of speculative scoping studies carried out across the boundaries of electronics and other disciplines within ICT. Some will be successful while others will show that further progress is not possible. In this way, the ICT community will be in a strong position to identify emerging research topics with economic and social impact at an early stage. Therefore, the ICT research community itself will be a beneficiary of eFuturesXD. The cross-disciplinary account will also be of benefit to those involved in the public sector forming future policy for funding research, which can be assessed in the light of this evidence. It will give credibility to the ICT community to lobby for greater research support in areas defined by the most successful scoping studies. The scoping studies will in some cases identify synergies between existing disciplines and lead to a closer cooperation between these experts. This may then lead to commercialisation of established research strengths in hitherto unforeseen areas of application to the benefit of the UK economy. This could be through existing UK companies who can readily exploit the research or through the launch of spin-out companies. In other cases totally new areas of research will be uncovered to be explored. These areas are likely to be more risky and long term in nature, but with the potential of a higher economic return as there is a higher likelihood of a global lead. Increased recognition globally of electronics research in the UK will lead to further inward investment contributing additional wealth creation.
Impact will also be in evidence in the people pipeline. Cross disciplinary research will improve the skill set and mobility of researchers, particularly those involved in the execution of scoping studies or having extended visits to laboratories across a discipline boundary. Cross disciplinary research means learning new skills in a different area or applying existing skills in a new way. This will make researchers more aware of, and confident with, their transferable skills and so of greater value to the ICT Community.

The potential to generate new knowledge at the boundary of disciplines is enormous. This will lead to a greater body of understanding generated from the successful new research that is anticipated to emerge from work initiated by eFuturesXD. Some of this will lead on to IP protection for future exploitation in the UK whilst in other cases the accumulation of know-how will be the more valuable commodity.

Publications

10 25 50
publication icon
Mokhov A (2014) Algebra of Parameterised Graphs in ACM Transactions on Embedded Computing Systems

publication icon
Mokhov A (2014) Synthesis of Processor Instruction Sets from High-Level ISA Specifications in IEEE Transactions on Computers

 
Description Piotr Dudek (Manchester). "Implementation of Cortical Models using Configurable Analogue Spiking Neural Network Circuits"
The project "A Tool Chain for running NeuroML/LEMS Models on the Bluehive" achieved a number of research results that contributed to the PhD thesis "Efficient programming models for neurocomputation". The expertise gained helped Prof Andrew Brown and Prof Simon Moore (collaborators on the project) to successfully bid for an EPSRC Programme Grant entitled "Event-based parallel computing - partially ordered event-triggered systems (POETS)" (EPSRC Reference: EP/N031768/1, value £4.98m).

Christofer Toumazou (Imperial College); Eleni Vasilaki (Sheffield) "Synaptic-like Dynamics of Volatile memristors".
Welcome Trust, 'The cortical representation of low-probability stimuli and its neuromorphic implementation', Fellow Dr Vanattou-Saïfoudine, collaboration with the institute for NeuroInformatics ETHZ/University of Zurich (Sep 2016 - Aug 2019). Supervisor.
The institute of neuroinformatics were co-authors in the article which was produced under futures funding: Berdan, R., Vasilaki, E. , Wei, S. L., Khiat, A., Indiveri, G., Serb, A. and Prodromakis, T. (2016), Emulating short-term synaptic dynamics with memristive devices. Scientific Reports, Nature Publishing Group, 6, 18639; doi:10.1038/srep18639

Liam McDaid (Ulster); John Marsland (Liverpool) "Hardware / Software models of brain like self-repair"
A far as we at Ulster and Liverpool are concerned the eFuturesXD funded project was extremely successful. The project produced an IEEE journal paper which demonstrated that the interactions between astrocytes and neurons holds the key to a new robust computational algorithm which could potentially led to a novel hardware self-repairing computing platform. Based on this work we approached staff at the University of York (Andy Tyrell's group) and after demonstrating the output of our eFuturesXD project we felt that a strong collaboration with a good case could be made to EPSRC where the objective of the bid would be to showcase our self-repairing algorithm using a range of robot applications, already existing at York, as demonstrators. This bid was successful and indeed was ranked first by the EPSRC the panel: the project is currently underway and producing promising results. The strength in this bid came from the work carried out within the eFuturesXD project. Furthermore, based on the collaboration to-date we (with York) are now submitting a phase 1 proposal to the Centre for Defense Enterprise (CDE) to bid for a small amount of funding to demonstrate our self-repairing algorithm as a controller to oversee robot navigation in a hazarders environment. If we are successful and deliver on phase 1 then we can potentially go for phase 2 funding from CDE which is up to £1 million.

Alex Yakovlev (Newcastle) "Hardware/Software Mechanisms for Cross-Layer Power Proportionality (PowerProp)"

Roger Woods (Belfast) "WifiEar: A new form of telecoil"
The funding provided an opportunity for us to investigate using embedded technology to investigate linking hearing aids to WiFi technology and improve their functionality. The grant led to the development of a research proposal with the University of Liverpool involving the Institute of Hearing Research and On-Semi to the EPSRC call on Transforming approaches to improving hearing aid technology in September 2014. Unfortunately, it was not successful but there are plans to resubmit. The work was also the subject of a submission called WiFiEar to the Northern Ireland Science Park Invent competition involved in commercializing the research. We won one of the runner-up competitions.

Timothy Constandinou (Imperial College) "Next Generation Neural Probes for Large-Scale Recording in the Living Brain"
As detailed in the final report, this award enabled a larger grant (EPSRC iPROBE, EP/K015060/1, £367k) that was recently completed. In addition to academic outputs - e.g. papers (which are very much still work in progress), we are also considering commercial exploitation (been exploring landscape for licensing technology developed to companies working on ePhys). Initiated this during the FENS Forum in Neuroscience earlier this year (July 16, in Copenhagen) where we approached a number of companies who all expressed interest. We have started more extensive discussions with one key player in the field, together with our technology transfer office- still in "discovery" phase, and establishing a memorandum of understanding. Another "followup development" is I have been awarded an EPSRC Early-Career Fellowship (2015-2020, "Empowering Next Generation Implantable Neural Interfaces", EP/M020975/1, £1m) that in part, uses the know-how acquired during the previous activities (eFuturesXD, EPSRC iPROBE project).

Simon Moore (Cambridge) "A toolchain for running NeuroML/LEMS models on the Beehive"

Peter Andras (Newcastle) "Neural System Functionality"
Since the project ended we had one related paper accepted:
Luo, L, Nikolic, P, Evans, B, Andras, P, Yakovlev, A, Degenaar, P (2016). Optogenetics in silicon: a neural processor for predicting optically active neural networks. IEEE Transactions on Biomedical Circuits and Systems, in press - available online, DOI: 10.1109/TBCAS.2016.2571339.

Thomas Nowotny (Sussex) "Fast electronic noses through spiking neuromorphic networks"
The findings from the eFutures funded project were initially presented at the Computational Neuroscience Society conference (CNS 2014, Diamond, Alan, Schmuker, Michael, Berna, Amalia Z, Trowell, Stephen and Nowotny, Thomas (2014) Classifying chemical sensor data using GPU-accelerated bio-mimetic neuronal networks based on the insect olfactory system. In: CNS 2014, 26-31 July 2014, Québec City, Canada.) and subsequently published fully in the journal Bioinspiration and Biomimetics (Diamond A, Schmuker M, Nowonty T., Classifying continuous, real-time e-nose sensor data using a bio-inspired spiking network modelled on the insect olfactory system, 2016, Bioinspiration & Biomimetics , doi:10.1088/1748-3190/11/2/026002 ). The spiking neural network model developed to classify odors was subsequently used in the well-known Human Brain Project in sub-project 9 (Neuromorphic Computing) as a benchmark model to compare implementations on three neuromorphic platforms. This comparison was presented at the INCF Congress on Neuroinformatics and the findings have been published fully in the journal Frontiers In Neuroscience (Diamond A, Nowonty T., Schmuker M., Comparing Neuromorphic Solutions in Action: Implementing a Bio-Inspired Solution to a Benchmark Classification Task on Three Parallel-Computing Platforms, 2016, Frontiers In Neuroscience, doi:10.3389/fnins.2015.00491 ) .

Anne Bernassau (Glasgow) "Acoustic Control and Digital Counting (ACDC)"

Tao Geng (Middlesex) "A Neuromorphic Controller for Stable and Adaptive Walking of Bipedal (two-legged) Robots"
The eFuturesXD project in neuromorphic robotics has greatly boosted our research. Based on the cooperation and work done in that project, we wrote a proposal titled "A Neuromorphic Control System for Agile Biped Walking", was funded by EPSRC (EP/P00542X/1, £561K).
Exploitation Route Research grants obtained and papers published (see above)
Sectors Electronics

 
Description eFuturesXD - crossing the boundaries
Amount £578,602 (GBP)
Funding ID EP/I038357/1 
Organisation Engineering and Physical Sciences Research Council (EPSRC) 
Sector Public
Country United Kingdom
Start 11/2011 
End 05/2015