Hyundai Motors: PHM (Prognostics and Health Management): Data-driven verification criteria of automotive electronics

The automotive industry is transitioning toward electrification and autonomous driving, making Software-Defined Vehicles (SDVs) essential. These new functions require high-performance electronic control units, while hardware complexity continues to increase, and development schedules become shorter.

To address this, we build an automated verification system and define verification criteria for harsh automotive environments. However, traditional verification rules are typically based on worst-case assumptions, leading to excessive margins and inefficiencies.

To improve this, we focused on SDVs and their ability to utilize in-vehicle operational data. One example is Prognostics and Health Management (PHM), which evaluates system health during operation and predicts failures in advance. PHM data can also be applied during development to support more optimal design decisions.

We introduce a method for optimizing verification criteria in the automated verification system using PHM data.