Predictive and Prescriptive Maintenance in the Context of Automotive Functional Safety

Abstract

The ever-changing landscape of advanced SOCs reshape traditional approaches of automotive functional safety (FuSa). Electrification (EV), connectivity, driver-assistance (ADAS), and software-defined vehicles (SDV) have ushered in the era of mega-functionality and scale. This paper discusses the paradigm shifts and required methodologies to navigate the surge of innovation and ensure the utmost performance, reliability, and safety of automotive systems.

Gain valuable insights into the implications of data-driven and transformative advancements in artificial intelligence, machine learning, and data analytics for automotive functional safety. Discover proactive approaches that transcend reactive measures, enabling stakeholders to effectively detect anomalies, anticipate failures, and proactively mitigate risks. By embracing continuous monitoring and predictive insights, automotive manufacturers and OEMs can achieve unprecedented levels of resilience, robustness, and operational efficiency, from ICs to ECUs.

This paper describes:

• How hardware failures can be anticipated by prognostics techniques
• How defects innate to manufacturing or caused by operational conditions and usage occur affect the lifetime of the system
• Time-To-Failure (TTF) predictions in the context of estimating the Failure Rate (FR) and the subsequent improvement of the reliability of an electronic device
• How deep data on-chip monitors, degradation modeling based on Physics-of-Failure, and analytics enable monitoring of the margin degradation of an IC
• How to estimate remaining useful life (RUL) and prevent future failures