Hardware, AI Software Collaboration Improves EV Battery Performance

Infineon Technologies has partnered with U.K.-based Eatron Technologies to allow the two companies to implement advanced machine learning algorithms on Infineon’s AURIX TC4x microcontroller (MCU). Working within an EV battery management system (BMS), powerful, predictive algorithms are designed to enhance EV battery performance and operating life.

 

Analytics improve EV range, charge times, and battery health. Image used courtesy of Eatron

 

The combination of Infineon’s processing hardware and Eatron’s AI-driven, advanced battery management software will allow the two companies to offer solutions that address key challenges related to greater EV adoption – vehicle range, battery charge times, and battery health.

Through the partnership, Eatron and Infineon will offer e-OEMs and Tier 1 suppliers a turn-key battery management solution that is pre-packaged, fully integrated, and tested with the most advanced hardware and software technologies.    

 

AI-Powered Battery Management Software 

For an EV, the battery pack is the core power source for the vehicle and its operation (such as charge and discharge) and is controlled by the software-drive battery management system.  

 

Eatron battery management software solutions. Image used courtesy of Eatron
 

With its intelligent battery management software, Eatron Technologies is pushing the boundaries of EV battery performance. The company’s AI-based solutions offer critical EV battery diagnostic capabilities like lithium plating detection and predictive capabilities that assess a battery’s state of health (SOH),  aging trajectory, and remaining useful life (RUL).  

According to Umut Genc, CEO of Eatron, combining Eatron’s predictive algorithms with Infineon’s advanced processing hardware will offer customers a leading BMS solution that provides accurate, cell-level assessments of a battery’s available charge, power, and long-term health prognosis.  

 

Battery state of health (SOH) and predictive analytics. Image used courtesy of Eatron

 

In addition to edge computing capabilities, Eatron’s BMSTAR automotive-grade BMS also incorporates connectivity capabilities that allow it to transmit battery data to the cloud for more advanced predictive analysis of battery RUL and fleet-wide performance trends.  

 

Cloud analytics for EV battery performance. Image used courtesy of Eatron

 

Advanced Microcontroller Capabilities

The AURIX TC4x family of MCUs incorporates advanced machine learning capabilities, including an integrated parallel processing unit (PPU) that will allow Eatron and others to implement their sophisticated AI-powered software algorithms effectively.  

 

AURIX TC4x MCU chips. Image used courtesy of Infineon

 

The PPU is a single instruction, multiple data vector digital signal processor that, according to Infineon, greatly reduces data processing times compared with traditional approaches.

The PPU is part of the MCU family’s AURIX Accelerator suite that also includes:

• Data Routing Engine (DRE) for efficient communication and data handling
• cDSP for digital signal processing for ADC signals
• Single Processing Unit (SPU) for radar accelerator
• Security Accelerators (CSRM/CSS) for hardware crypto acceleration

Each AURIX TC4x  MCU has six TriCore CPUs that operate at speeds up to 500 MHz with a 1.8 V core.

 

AURIX TC4x MCU architecture. Image used courtesy of Infineon

 

According to Thomas Boehm, senior vice president and general manager of Microcontroller Automotive at Infineon, the AURIX TC4x addresses the limitations of prior generations of processing solutions and is designed to improve modern xEV performance across a range of vehicle applications, including predictive and diagnostic analytics in powertrain battery systems.