AI model uses smartphone location data to predict power grid usage

In a paper published on the preprint server Arxiv.org, Microsoft and the University of Washington researchers propose an AI system that uses smartphone location data to forecast electrical load. They say their architecture, which takes into account data from geographical regions both within the U.S. and Europe, can outperform conventional forecasting methods by more than three times. The pandemic shows no sign of abating, and it’s made a striking impact on the global electrical grid. Stay-at-home orders and social distancing meant to slow the outbreak of COVID-19 have resulted in major shifts in load patterns and peak demands. Italy saw a 25% reduction in demand during its country-wide lockdown, and in the U.S., overall power consumption has fallen to a 16-year low.