SPANNER Workshop
Self-repairing Hardware Paradigms based on Astrocyte-neuron Models
3rd September 2018, York, UK


The final programme is available here

The SPANNER Workshop on Self-repairing Hardware Paradigms based on Astrocyte-neuron Models will be held in the historic city of York. The workshop aims to highlight and discuss emerging trends and future directions in the fields of bio-inspired fault tolerant systems, particularly self-repairing spiking neural networks, with applications in autonomous robotic systems and beyond. The workshop will feature invited position papers from leading researchers.

The technical programme will focus on the potential for future developments within the field of bio-inspired fault tolerant systems, addressing areas including (but not limited to):

  • Astrocyte-neuron interactions and modelling
  • Spiking neural network models, algorithms, and hardware implementations
  • Training, learning, and optimisation in spiking neural networks
  • Biological models for fault tolerant systems
  • Implementation of bio-inspired and fault tolerant systems on FPGAs
  • Hardware optimisations for bio-inspired systems
  • Spiking neural network and bio-inspired controllers for autonomous robotics
  • Autonomous swarm robotic systems
  • Fault tolerance in robotic systems
  • Innovative design techniques for bio-inspired and fault tolerant systems

Ana Covelo FernndezFunctional consequences of neuron-astrocyte communication
Ana Covelo Fernndez is a Postdoctoral researcher at the Department of Neuroscience, University of Minnesota, USA.
Simon DavidsonSpiNNaker-2
Abstract. SpiNNaker was designed to be a flexible, low-power neuromorphic computing platform allowing both computation neuroscientists and roboticists to design large-scale neural networks running in real time. Since the first SpiNNaker chips became available in 2011, we have built up a user base of more than one hundred institutions around the world. Now, as part of the Human Brain Project (HBP) we - together with partners in TU Dresden - have been working on the second generation SpiNNaker architecture. In my talk I'll present this new architecture and discuss how we got there, including the lessons we learned from the first SpiNNaker machine and our plans for SpiNNaker going forward.
Simon Davidson is a Research Fellow in the SpiNNaker project at the Advanced Processor Technologies group, School of Computer Science, University of Manchester, UK.
Jennifer HaslerPhysical Computing towards building Large-Scale Brain Computations
Abstract. Cognitive Neuromorphic systems are gaining increasing importance in an era where CMOS digital computing techniques are meeting hard physical limits. These silicon systems mimic extremely energy efficient neural computing structures, potentially both for solving engineering applications as well as understanding neural computation.
Neuromorphic techniques are of increasing interest along with other Physical computing directions, such as Analog, Quantum, and Optical computation. Understanding and developing computational theory of physical computation became relevant with the advent of large-scale Field Programmable Analog Arrays (FPAA) as well as other recent physical computing implementations. Digital computation is enabled by a framework developed over the last 80 years. Analog computing techniques result in 1000x improvement in power or energy efficiency, and a 100x improvement in area efficiency, compared to digital computation. FPAA structures have demonstrated applications examples range from acoustics and sensor processing, classification, embedded machine learning, image processing, communications, RF signal processing, and optimal path planning using coupled PDEs.
Towards this end, we provide a glimpse at what the technology evolution roadmap looks like for these systems so that Neuromorphic engineers may gain the same benefit of anticipation and foresight that IC designers gained from Moore's law many years ago. Scaling of energy efficiency, performance, and size will be discussed as well as how the implementation and application space of Neuromorphic systems are expected to evolve over time. We will focus on a number of techniques for long-term modulation processes applicable to learning mechanisms as well as other slow modulatory processes.
Jennifer Hasler is a Professor at the School of Electrical and Computer Engineering, Georgia Institute of Technology, USA.
Alan WinfieldExperiments in Artificial Theory of Mind
Abstract. Theory of mind is the term given by philosophers and psychologists for the ability to form a predictive model of self and others. In this talk I will advance the hypothesis that simulation-based internal models offer a powerful and realisable, theory-driven basis for artificial theory of mind. Proposed as a computational model of the simulation theory of mind, our simulation-based internal model equips a robot with an internal model of itself and its environment, including other dynamic actors, which can test (i.e. simulate) the robot's next possible actions and hence anticipate the likely consequences of those actions both for itself and others. I will present a series of experiments which each demonstrate some aspect of artificial theory of mind.
Alan Winfield is a Professor of Robot Ethics at the Department of Engineering Design and Mathematics, University of the West of England, UK, and conducts research within the Bristol Robotics Laboratory.
Karlheinz MeierHow to compute with physical laws
Abstract. The success of modern computing is entirely based on the idea of describing parameters by abstract symbols and function by algorithms acting on the parameters. The Turing machine is a corresponding complete mathematical theory of computation. Several specific architectures like the von Neumann architecture and its many variations have been devised to implement this approach with outstanding successes.

Many current challenges of computing address the simulation of complex systems, in which a very large number of elements interact with each other. Typically, the number of elements is much larger than the number of processors and the issue of scaling is of crucial importance. Scalability issues are of particular importance when the computational goal is to describe the dynamics of the complex system at different time scales.

The brain is a particular example of a complex system with timescales covering milliseconds to years corresponding to more than 10 orders of magnitude. The mechanisms of learning and plasticity at those scales are of extreme importance for advancing brain science but equally important for the development of a more biologically grounded AI with a more generic application space.

Physical model neuromorphic systems using analog or mixed-signal circuits emulating neuronal dynamics have been proposed and built to overcome the scaling and simulation time problem in particular for the implementation of learning processes. In my talk I will introduce the idea of physical computing, show actual implementations and applications with special emphasis the exploitation of stochastic processes and very recent ideas for biological approaches to deep learning.
Karlheinz Meier is a Professor of Experimental Physics at the Department of Physics and Astronomy, University of Heidelberg, Germany, and Director of neuromorphic computing for the EU Human Brain Project.
Eleni VasilakiSynaptic plasticity and learning in spiking neural networks.
Abstract. Neuroscience has long been an inspiration for Artificial Intelligence and Machine Learning. In this talk I will present some fundamental ideas about biological learning, and how these are related to Machine Learning. More particularly I will discuss how simple concepts from the 1940's underline ideas about memory formation in both artificial and natural systems. I will then proceed to recent experimental and theoretical results and discuss what do they tell us about the way information is encoded in different brain areas. I will also show how these results not only inspire Artificial Intelligence but also future electronics, and present concrete examples where novel electronic circuits demonstrate functionality akin to biological circuits.
Eleni Vasilaki is Professor of Computational Neuroscience & Neural Engineering and Head of Machine Learning Group, University of Sheffield.

Exhibition Centre, University of York

The workshop will be held at the University of York's Exhibition Centre, located on Campus West. The Exhibition Centre lies at the heart of the University, next to Central Hall, providing a unique and modern venue just a short distance from the city of York. Click here for directions.

The City of York

Internationally acclaimed for its rich heritage and historic architecture, York's bustling streets are filled with visitors from all over the world. Within its medieval walls you will find the iconic gothic Minster, Clifford's Tower and the Shambles - just a few of the many attractions.

York boasts specialist and unique boutiques but also all the high street stores on its busy shopping streets. Alongside them you will find cinemas, theatres, an opera house, art galleries, a vast range of restaurants, live music venues and clubs. York is particularly renowned for its multitude of pubs and bars, from the modern to the medieval.

See for more information.

Travelling by air

Manchester Airport is a large airport in the north of England, and has a wide range of international flights and connections via London. Trains run directly to York from the airport station and take just under 2 hours. This is generally the most convenient option.

London Heathrow is the largest UK airport, with flights to a wide range of international destinations. Upon arrival, take the Heathrow Express train to Paddington station, then change to the Hammersmith and City underground line to reach King's Cross station (this takes about 30-45 minutes). Direct trains run frequently to York and take about 2 hours. London Gatwick, London Stansted, and London Luton also have public transport connections to York.

Leeds-Bradford is the closest airport to York, and has some international flights. Taxis to York take around 45 minutes. Other nearby airports with public transport connections include Newcastle, Durham Tees Valley and Humberside.

Travelling by rail

From Europe — York can be reached in around 5 hours from Paris or Brussels by train, by taking the Eurostar from Paris Nord to London St Pancras, with a short transfer (5 minute walk) to London Kings Cross for a direct rail service to York.

From the United Kingdom — York is on the East Coast main line from London to Edinburgh, just over two hours away from London King's Cross and around 2.5 hours from Edinburgh. There are also direct express services to many other major cities, including Manchester, Newcastle, Sheffield, Leeds, Birmingham and Glasgow.

Further travel details can be found here.

On Campus

We have a limited number of single en-suite rooms available in James College, which is only a 2-minute walk from the Exhibition Centre where the workshop will be held.

Priority will be given to authors of accepted contributions. We also have a number of student bursaries available, which will provide free accommodation and contribute towards the travel costs of registered PhD students. If you would like to be considered for free accommodation, please indicate this in your registration email.

Off Campus

York has a wide selection of hotels and B&Bs. However, as a major tourist destination, it is recommended that you book accommodation early. The University of York is approximately 2 miles from the city centre, and is served by a regular bus service.

These hotels all lie close to the bus route to the university, and are also convenient for the city centre

Novotel York Centre.

Park Inn by Radisson.

Holiday Inn York City Centre.

Hotel Indigo York.

The York tourist office provides an accommodation search facility.

Invited speakers

Ana Covelo Fernndez
University of Minnesota, USA

Simon Davidson
University of Manchester, UK

Jennifer Hasler
Georgia Tech, USA

Alan Winfield
Bristol Robotics Laboratory, UK

Karlheinz Meier
University of Heidelberg, Germany

Eleni Vasilaki
University of Sheffield, UK


David Halliday

Andy Tyrrell

Jon Timmis

Alan Millard

Anju Johnson

Department of Electronic Engineering,
University of York

Liam McDaid

Jim Harkin

Junxiu Liu

Shvan Karim

School of Computing,
Engineering and Intelligent Systems,
Ulster University