Research Article

Machine Learning and Safety Standards in Autonomous Vehicle Systems: A Technical Overview

Authors

  • Rajani Acharya University of Southern California, USA

Abstract

This article examines the integration of machine learning (ML) algorithms with safety standards in autonomous vehicle (AV) systems, with a focus on modern perception systems, advanced ML implementations, and evolving compliance frameworks. It explores how modern autonomous vehicles leverage multi-modal sensor fusion and deep learning to enhance system reliability while maintaining strict safety standards, resulting in high performance perception systems with 92% object detection accuracy at distances up to 120 meters and 94.5% classification accuracy through sensor fusion. The application of ISO 26262's fail-silent design, ASPICE process maturity, and UL 4600’s system-level validation ensures 91.2% requirements traceability and 99.7% fault detection coverage across safety-critical components while maintaining 99.95% system uptime during primary sensor failures. Technical optimizations through model quantization achieve further reduction in computational requirements and improved accuracies. Through analysis of current implementations and emerging technologies, this study presents a comprehensive technical overview of the integration of ML and safety standards in AVs, examines technical challenges such as real-time processing and bias mitigation, and explores emerging research directions including transformer-based perception models and cross-domain generalization for future-safe autonomous systems.

Article information

Journal

Journal of Computer Science and Technology Studies

Volume (Issue)

7 (3)

Pages

851-859

Published

2025-05-23

How to Cite

Rajani Acharya. (2025). Machine Learning and Safety Standards in Autonomous Vehicle Systems: A Technical Overview. Journal of Computer Science and Technology Studies, 7(3), 851-859. https://doi.org/10.32996/jcsts.2025.7.3.94

Downloads

Views

42

Downloads

46

Keywords:

Autonomous Vehicles, Machine Learning, Safety Standards, ASPICE Compliance, Perception Systems