Dr Mohammad Amir

One Board, Many Devices: making wearable technology quicker to build

Dr Mohammad Amir Pillar 2 - Low-power semiconductors and digital design Dr Mohammad Amir, a REACT researcher in Pillar 2 - the group of Prof. Hadi Heidari, discusses in his recent REACT interview with Jack Marley McIntyre, his development of an ultra-low energy modular electronics platform for wearable and sensor-rich devices. "The idea is simple: separate the parts every device need from the parts that make each device different. The shared part is a small core board carrying the processor, the wireless link, the battery and power management, and the software that ties them together. It is designed once, tested properly, then reused." "The part that changes is a second board that plugs onto it, carrying whichever sensors an application needs, such as heart or brain signals, movement, temperature or air quality. It works rather like changing the lens on a camera instead of buying a new camera." Q: But why is a design like this important? "A smartwatch that counts your steps, a patch that keeps an eye on a patient's heart at home, a sensor that measures air quality in a classroom: these devices look nothing alike but open them up and you find much the same ingredients. A tiny computer, a wireless radio, a battery and a handful of sensors. Yet almost every time someone sets out to build one, the electronics are designed again from a blank sheet. That repetition is the problem my research is trying to solve." Q: You've identified repetition as the area you are trying to impact with your research, why is repetition in R&D a negative? And how do you see your board having an impact on the electronics sector? "Designing custom electronics is slow and expensive. A single redesign can take months, and each new board brings a fresh chance of mistakes, delays and cost. For a research group testing an idea, or a small company trying to reach a first product, that is often the difference between a project going ahead and quietly stopping. Reusing a proven core lets teams spend their time on the question that matters does this device help the person wearing it? rather than on rebuilding the same foundations." "How I see this platform having an impact, is that it is also built for very low power use from the outset, which is what makes small battery-powered devices practical rather than frustrating." "Because the core is reused rather than thrown away with each iteration, the approach is kinder to budgets and produces less electronic waste. And it is designed for devices that interpret their own data on board rather than sending everything to the cloud, which keeps personal information private, saves battery life, and works where there is no reliable wireless connection." Dr Amir then takes out a small 2 to 3 inch PCB, and effortlessly connects it to an input. Dr Amir, Presenting his R&D findings to the REACT partnership Q: So now that you've designed and manufactured this prototype, what is the next stage for this device and your R&D? "This is the first prototype (ed. while giving a demo of the platform), and as you can see, it is built and ready to use. Next comes measuring how little power it really uses, checking how reliably it sends data, and adding the first sensor boards." "The following version will move to a processor that combines several different kinds of computing core, which will enhance the edge-AI applications the platform can support by letting the device interpret far more of what it measures on board, using the most efficient core for each task rather than one core for everything." "Managing the power budget as that capability grows will be a key part of the work." Q: What sectors or usage cases do you see your low power, edge-AI platform having the biggest impact? "First of all, Health researchers and clinical teams building monitoring devices for trials or for use at home. I am also keen for this prototype to support small companies and start-ups with a strong product idea but no in-house hardware team. In REACT we have also identified environmental and industrial monitoring projects needing many small, battery-powered wireless sensors which is an exciting usage case. Lastly from experience, Universities and students, who can build on a tested base instead of starting from nothing." Q: What would be your ask to the wider industry in relation to your work? "I would be glad to hear from anyone with a sensing problem worth solving, particularly wearable hardware developers, sensor integrators and low-power specialists." If you are looking for sustainable, ultra-low power edge-AI developments in electronics and beyond, explore how REACT could help by visiting our website https://www.react.ac.uk or by contacting Dr Mohammad Amir at mohammad.amir@glasgow.ac.uk.