Certifiability Analysis of Machine Learning Systems for Low-Risk Automotive Applications

IF 2 4区计算机科学 Q3 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Computer Pub Date : 2024-08-30 DOI:10.1109/mc.2024.3401402

Vinod Vasudevan, Amr Abdullatif, Sohag Kabir, Felician Campean

引用次数: 0

Abstract

We analyze the existing safety standard, ISO 26262, for automotive applications, determining the certifiability of machine learning (ML) approaches used in low-risk automotive applications. This can help assuring the security and safety of ML-based autonomous driving systems, gaining the trust of regulators, certification agencies, and stakeholders.

查看原文本刊更多论文

低风险汽车应用中机器学习系统的可认证性分析

我们分析了适用于汽车应用的现有安全标准 ISO 26262，确定了低风险汽车应用中使用的机器学习（ML）方法的可认证性。这有助于确保基于 ML 的自动驾驶系统的安全性，赢得监管机构、认证机构和利益相关者的信任。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Computer 工程技术-计算机：软件工程

CiteScore

3.90

自引率

4.50%

发文量

378

审稿时长

12 months

期刊介绍： Computer, the flagship publication of the IEEE Computer Society, publishes peer-reviewed articles written for and by computer researchers and practitioners representing the full spectrum of computing and information technology, from hardware to software and from emerging research to new applications. The aim is to provide more technical substance than trade magazines and more practical ideas than research journals. Computer seeks to deliver useful information for all computing professionals and students, including computer scientists, engineers, and practitioners of all levels.