German automotive supplier ZF has introduced TempAI, an artificial intelligence-based software solution that enhances temperature prediction and control in electric vehicle (EV) drives, aiming to significantly improve both efficiency and performance.
The system relies on a self-learning model to predict internal motor temperatures with over 15% more accuracy, allowing for precise thermal management without requiring additional sensors or hardware. By leveraging only existing control units and low computing resources, TempAI presents a cost-efficient and scalable approach for series production.
“With TempAI, we can further increase the efficiency and reliability of our drives,” said Dr. Stefan Sicklinger, Head of AI, Digital Engineering, and Validation in R&D at ZF. “This technology demonstrates how data-driven development can be not only faster, but also more sustainable and more powerful.”
Due to the inability to measure temperature accurately within the motor’s core, manufacturers have traditionally relied on conservative cooling strategies to avoid overheating. This has often led to oversized or overused cooling systems, additional material use, or underutilized performance capacity.
In contrast, TempAI enables real-time, high-precision thermal utilisation of electric motors, allowing operation closer to their thermal limits. ZF reports that the improved temperature forecasting results in up to 6% more peak power, with energy consumption reductions of 6–18% under dynamic conditions, including during demanding test laps like the Nürburgring Nordschleife.
The software is also expected to streamline EV powertrain development. ZF claims that design time for thermal systems can be reduced from months to days, while also enabling lower use of heavy rare earth materials due to optimized cooling requirements.
The company confirmed that TempAI is now ready for series production and will be available with ZF’s latest generation of electric motors, including the recently launched SELECT drive platform.