How ZF Uses AI to Make Electric Motors More Efficient

The temperature management of electric motors is an enormously important factor in electromobility. In order to optimise operating strategies during the development of e-motors and to allow the electric machines to run more efficiently and with higher continuous power during operation, precise determination of temperatures is essential. Supplier ZF introduces TempAI, a method of determination that uses smart algorithms to venture into areas where conventional methods are simply impossible, too hot, or too tight.

TempAI is a temperature model that automatically generates physically based models from measurement data and makes them operational in the shortest possible time. Existing control units are sufficient for this, according to ZF, no additional hardware is required. The AI models used are also said to require low computing resources.

Rheinmetall Charging Kerb Ready for Regular Operation

After just over a year, Rheinmetall's charging kerb pilot project in Cologne has been successfully completed, proving the market readiness of the solution: Technical availability during the test period was over 99%, and users consistently gave positive feedback.

Dr. Martin Large

Intelligent thermal management is an important topic for the supplier. ZF recently presented a new system called TherMaS, which is intended to efficiently and compactly temper the battery, e-machine, power electronics, and interior. The Friedrichshafen-based company relies on a modular, technology-open, and systemic strategy in the field of e-drives, with the new so-called Select platform intended to help combine costs, variety, and speed.

Advantages in Operation and Development

At ZF, they talk about a 15 percent more accurate temperature control through Artificial Intelligence. "TempAI is a real technological breakthrough for the temperature management of electric drives," emphasises Otmar Scharrer, Head of Development for Electrified Drive Technologies at ZF. According to ZF, the more precise temperature prediction enables more targeted control up to the thermal operating limit. The result here: up to six percent more peak performance and a verifiable increase in efficiency in the WLTP cycle. During dynamic driving - such as on the Nürburgring Nordschleife - energy consumption is said to decrease by six to 18 percent depending on the load point, according to the supplier.

ZF sees advantages particularly in the development of electric drives. There, AI helps to understand and capture processes inside the electric motor for which there is no physically reliable model due to cost or time reasons. "This technology allows us to further increase the efficiency and reliability of our drives. At the same time, with TempAI, we demonstrate how data-driven development can be not only faster but also more sustainable and powerful," says Stefan Sicklinger, Head of AI, Digital Engineering and Validation in the R&D department.

Among the explicit challenges is measuring the temperature inside a rotor. This can only be directly measured during operation at high cost. Here, measurement data come into play, which result from elaborate functional tests on the test bench and are later systematically recorded in the test vehicles, such as temperature values from the environment, for example from the oil pan, as well as the rotor speeds. From the various possible operating points and their temporal progression, millions of data points arise. Accordingly "trained", AI algorithms would filter out exactly those dependencies that are particularly significant for the temperature changes on the rotor and stator, according to ZF.

Much shorter development and fewer rare earths

Through the optimised thermal design, significant amounts of heavy rare earths can also be saved. At the same time, the development time per project is significantly reduced: according to ZF, from several months to a few days. The AI-based technology is ready for series production and available for the new generation of ZF electric motors.

This article was first published at automotive.eu 

automotiveIT car.summit

SV Tower 11 November 2025, Munich