The NWO has awarded two Venis to promising young EEMCS researchers, Jochen Cremer and Sijun Du. Jochen Cremer will use his grant to research how AI can help to reduce power failures caused by the energy transition and the consequent far-reaching digitization of our energy grid. A key role will be to improve the understanding of physical laws by Artificial Intelligence (AI). Sijun Du will investigate solutions for improving the sustainability of the 'Internet of Things' (IoT) – which is making our environment increasingly digitized. By allowing chips to harvest energy from their environment, small IoT computer systems can become self-sufficient.
Teaching physics to AI to avoid power outages
The energy grid of the future will use a complex network of small, sustainable energy sources, such as solar panels and wind turbines. Increased complexity will make the network vulnerable to disruptions, made still worse by the extreme weather events caused by far-reaching climate change. Sudden catastrophic power outages can take place that potentially last for months and span entire regions, with serious consequences for society. Effective countermeasures depend on understanding the causes of these blackouts quickly, and Jochen Cremers's research uses artificial intelligence both to predict power outages and to identify and address effective solutions. By managing these risks, the research will help to accelerate the energy transition and protect society from the next pan-European power outage.
A sustainable Internet-of-Things, harvesting its own energy
More and more of our devices are being connected to the internet, with major benefits across anything from healthcare or smart farming to domestic dwellings. For example, an internet-connected pacemaker could send a timely signal to doctors, and a thermostat could heat only the rooms you are using. The Internet of Things (IoT) would be significantly improved in the future if we could use wireless sensors that are self-sufficient and small, with energy harvesting capabilities and dimensions smaller than a millimeter. At present, self-sufficient devices are larger, with their dimensions determined by off-chip inductors. Removing those inductors could potentially reduce system size by 99% and pave the way for real-world implementations and commercialization. Sijun Du’s project proposes a new inductorless architecture for energy harvesting. It achieves higher energy efficiency than conventional systems, breaks the miniaturization barrier and enables fully silicon-integrated, self-sufficient wireless sensors for a future Internet-of-Things.