MARCH: Magnetic Architectures for Reservoir Computing Hardware

Classical digital computing is power hungry, fragile, and hard to interface to the analogue real world.

Unconventional computers such as in materio Reservoir Computers (RCs) can help overcome these issues, particularly by being able to perform embodied computation that can directly exploit the natural dynamics of their material composition, thereby dramatically reducing the power requirements. Such devices have been proven in principle, but systems need to be scaled up to provide sufficient and appropriate computational power for real world tasks. Furthermore, a range of device configurations are needed to support different computational tasks.

Our aim is to combine multiple material RCs:
1. with similar instance configurations, to provide scalability
2. with diverse instance configurations, to perform together tasks that no single RC configuration can readily perform alone.

Magnetic materials provide an excellent flexible testbed for developing a material RC design process. Such materials have the intrinsic memory and complex non-linear dynamics needed for RC operation,
and also have well-established methods of interfacing for data input/output, needed to build a practical device. We will exploit the properties of patterned 2D layouts of magnetic nanoring wires, which can be readily manufactured with existing technologies. This flexible design and experimental platform will allow us to develop generic techniques that will apply across a range of smart material RCs.

We will develop new multi-RC design tools, and will design, test and manufacture nanomagnetic RCs. This toolset and manufacture will provide a robust and reliable testbed on which we will develop and evaluate scalable RC architectures that can be configured for a range of practical computing tasks. We will demonstrate the resulting multi-reservoir architecture in an audio-controlled robot: a `turtle' following simple LOGO-like commands, controlled purely by in materio reservoirs, with no onboard digital computers.

The output will be a new design methodology and platform for multi-reservoir devices, that can be exploited to design low-power, robust, flexible, and efficient smart sensing and other `edge-computing' devices in a diverse range of materials.

Our aim is to use nanomagnetic systems as a robust and scalable test bed with which to explore how reservoir computers (RC) can be scaled-up towards use in challenging, real-world edge computing applications. Our primary focus is on generating of new knowledge. So in the short term (1-5 years) our research will impact the academic community through our research-enabling results, which we believe will inspire entirely new lines of enquiry.

To help us in engage with industry we have created an advisory board that includes industrialists and industry-facing academics. Their expertise will be invaluable in steering our project towards directions with genuine technological promise and helping us to develop industry contacts. We have planned an industrial workshop at the end of our project as an ideal forum for disseminating our ideas to industry. One of our advisory board members will validate our approach in a different material, contributing a major parallel project to integrate our MARCH design technologies with MEMS sensors, and use these to demonstrate advanced sensing and control systems in large industrial environments, within their Industry 4.0 setting. So we will have diverse range of proof-of-concept devices to act as powerful demonstrators of the potential of RC. We will be well-placed to engage industrial partners through established knowledge transfer frameworks and work collaboratively with them towards achieving higher technological readiness levels (5-10 years).

In the longer term (10+ years) economic impact will be generated as innovative technologies enabled by our research are realised and brought to market. There will also be substantial societal impact from these technologies; RC has the potential to revolutionise our ability to analyse and react to data `at the edge'; this will be hugely significant in areas as diverse as healthcare, structural health monitoring, manufacturing, robotics, security, and consumer electronics. Industrial engagement, through the mechanisms identified above, will be key in realising this impact.

Our project is highly interdisciplinary and will require team members with diverse skill sets. Project post-doctoral researchers will be trained in a range of cutting-edge techniques including device fabrication, materials characterisation, numerical modelling, microelectronic system design, and machine learning. As the programme is inherently collaborative in nature, the researchers will gain exposure and understanding of areas across the project, providing them with a diverse knowledge base that will be invaluable for future careers in academia or industry. The wider project will incorporate both PhD students and undergraduate student interns, allowing us to extend these important training opportunities to scientists at earlier stages of their careers. The impact of training will be achieved over the 3.5-year timescale of the project.

We have planned an extensive programme of outreach activities to ensure that our work will benefit the general public. We will work with a leading young artist to create an interactive art exhibit that allow the public to access the ideas behind RC. The exhibit will be shown at both STEM-based fairs and art-based events to ensure that people with a diverse range of interests are exposed to the ideas behind our project. Members of our team will also present our work in accessible terms at a variety of public exhibitions including Sheffield's Festival of the Mind, York Festival of Ideas, TEDx, Big Bang Fair, Cheltenham Science Fair, and the Royal Society Summer Science Exhibition. Our project website will include an area for non-technical readers, with short animated videos explaining the concepts that underpin our project, created in collaboration with local digital media companies. We also aim to write one article a year for popular science magazines/websites such as New Scientist, Scientific American, Science Daily and The Conversation.