Software

BMS Diagnostics Machine learning has transformed diagnostics by enabling real-time detection and prediction capabilities that were previously unattainable with conventional algorithms, leveraging increased compute capabilities onboard and in the cloud. It is now possible to detect li-plating, torn cell tabs, and predict risk of thermal runaway with real-time capable machine learning models in a BMS. MODERN HARDWARE ENABLING DEPLOYMENT ACROSS APPLICATIONS Infineon Aurix TC4 with PPU Accelerator

Achieves 95% speedup for ML features like lithium plating detection and remaining useful life RUL) prediction.

Syntiant Edge AI Accelerator

Supports a wide range of applications: Consumer electronics; E-bikes; Micro light commercial vehicles LCVs; Forklifts with low power ML acceleration capability.

REAL WORLD IMPACT

Software Updates Over Replacements Safety recalls are increasingly handled by diagnostic software updates rather than full battery replacements. Example: The Porsche Taycan recall

INNOVATIVE PARTNERSHIPS LINKED CLOUD AND EMBEDDED SOLUTIONS Edge and Cloud Compute to increase accuracy

Linked ecosystem approach to diagnostics, prognostics, SOH, and RUL

Adaptive BMS

Leverages online and offline compute capacity for batch analytics and increased accuracy, reduced false positives and negatives

LG Energy Solution LG ES & Qualcomm

LG ES diagnostics software deployed on Qualcomm automotive SoCs.

Leverages high compute capacity for improved diagnostics and performance.

Extracted from the PDF; charts and images may read imperfectly.