Abstract:

Power-on Self-Test (POST) is a cost-effective approach for detecting safety-critical faults in automotive systems, ensuring reliability and functional safety. This project introduces an enhanced multi-cycle logic Built-In Self-Test (BIST) technique that minimizes fault masking by leveraging sequential observation through a novel scan cell structure. Unlike conventional BIST methods, which may fail to detect faults over multiple test cycles, the proposed approach improves fault coverage by systematically observing and analyzing fault propagation across sequential states.Additionally, the system integrates adaptive fault diagnosis mechanisms to enhance real-time fault detection and mitigation. By incorporating machine learning-based anomaly detection, the system ensures early identification of potential failures, thereby improving overall vehicle safety and reliability. This method is particularly beneficial for advanced driver assistance systems (ADAS) and autonomous vehicles, where real-time functional safety assurance is crucial. The proposed solution meets industry standards such as ISO 26262 and can be effectively implemented in modern automotive electronic control units (ECUs), making it a robust and scalable approach for ensuring automotive functional safety.

Description:

.