The aim of the partnership is to develop advanced automotive battery management systems (BMS). Due to the state-of-the-art machine learning capabilities of the AURIX MCU family with its integrated parallel processing unit (PPU), Eatron will be able to maximise the performance and accuracy of its AI-powered battery management software.
The venture between the two companies will help electric vehicle (EV) manufacturers solve three technological challenges that have so far hindered increasing customer adoption of EVs: range anxiety, charging speed and battery health.
“Infineon’s AURIX TC4x PPU enables us to offer AI-based battery diagnostics, including lithium plating detection, as well as prognostics, such as state of health (SoH) and aging trajectory, and remaining useful life (RUL) prediction at the edge,” explained Umut Genc, CEO of Eatron. “Coupled with our state of everything (SoX) solution, which provides the most accurate and robust cell-level estimation of available charge, power, and battery health, this gives users of the TC4x the opportunity to have the leading BMS solution enabled by our software and machine learning models.”
The PPU, an on-chip single instruction, multiple data (SIMD) vector digital signal processor (DSP), has been designed to significantly reduce computation time compared to traditional CPUs. For ease of use, Infineon offers an automated toolchain within its ecosystem to ensure the most convenient and efficient handling of the PPU. For example, an automated toolchain helps customers to automatically convert existing models into vectorized code.
“Our new AURIX TC4x family is designed to improve the efficiency of many xEV applications. It overcomes previous limitations in computing performance, enabling our customers to fully exploit the value of high-precision models and algorithms,” said Thomas Boehm, Senior Vice President & General Manager, Microcontroller Automotive at Infineon. “The next level of technological advancement in electromobility is now also gaining momentum in BMS.”
