Glasgow-based semiconductor startup weeteq is developing Ultra Edge, a technology that uses lightweight machine learning on existing microcontrollers to predict electrical transients and respond in real time. It aims to reduce excess design margins, component stress, and energy use, without extra sensors, cloud infrastructure, or dedicated accelerators. Target areas include motor drives, industrial automation, automotive systems, electric vehicles, and AI data centers.
The team uses MATLAB, Simulink, and Simscape Electrical to model systems, test control strategies, and validate designs before moving to hardware. weeteq has demonstrated the technology on a data center prototype, received Best Paper recognition in the AI/HPC category at the OCP EMEA Summit, and participated in the Nissan-led SUITE electric vehicle efficiency project, and is now working toward real-world deployment. Its approach explores how simulation and AI can support power-efficient engineering.
Image generated by: ChatGPT
