没有合适的资源?快使用搜索试试~ 我知道了~
资源推荐
资源详情
资源评论
Submitted by the representatives of Canada, EC, USA, OICA Informal document WP.29-193-20
193rd WP.29, 25-28 June 2024
Agenda item 2.3
Considerations on Artificial Intelligence in the context of
road vehicles
This document is based on ECE/TRANS/WP.29/2024/34/Rev.1.
The Contracting Parties to the 1958, 1997 and the 1998 Agreements, participating in the
World Forum for Harmonization of Vehicle Regulations:
Having recognized the significant penetration of some Artificial Intelligence (AI) in
wheeled vehicles covered in the scope of the agreements administered by the World Forum
for Harmonization of Vehicle Regulations (WP.29),
Having noted that industry currently could use machine-learning tools to support the
development and/or testing of software before deployment,
Having discussed the technical fundamental aspects of some Machine Learning
systems in automotive products, to which the general public refers to as Artificial
Intelligence, and discussed corresponding definitions (see Annex 1),
Recalling the adoption of Recommendations on uniform provisions concerning cyber
security and software updates,
Having recognized the importance of also taking societal impacts into account during
the implementation and application of AI technologies,
Having assessed the importance of proper AI lifecycles for compatibility with existing
certification regimes,
Having reviewed existing AI use cases, to date, provided by Industry (see Annex 2),
Having acknowledged that the use of such technology for automotive applications is still
under development,
Having considered input from Industry on the potential interactions of AI with the New
Assessment Test Methods (see Annex 3) developed by the Working Party on
Automated/Autonomous and Connected Vehicles, as an example,
Have agreed on the following considerations in relation to using AI-based algorithms within
their automotive products:
Software
1. When implementing AI based algorithms and software, manufacturers, suppliers and
AI developers should take into account safety, security, environmental performance,
societal impacts and other relevant requirements.
Software update
2. This consideration document applies to certification requirements and Conformity of
Production. Any automated software update coming directly out of AI functionality,
including AI based algorithms and software updates issued by manufacturers, suppliers and
AI developers, which will significantly modify already certified functions according to the
recommendations on uniform provisions concerning cyber security and software updates
without resuming the relevant certification procedure, shall not be permitted.
3. Considerations must be given that after having trained an AI-system which is
incorporated in the software it should be validated by authorised parties and or certification
processes and assessed with regards to safety, security and environmental performance, and
other relevant requirements. Non-Certified systems containing AI, shall not influence
certified systems in a way it harms the certification. Following that process, the validated
software may be deployed in vehicles of a vehicle type. Data to be used for AI based system
development.
资源评论
SafetyResearch
- 粉丝: 0
- 资源: 3
上传资源 快速赚钱
- 我的内容管理 展开
- 我的资源 快来上传第一个资源
- 我的收益 登录查看自己的收益
- 我的积分 登录查看自己的积分
- 我的C币 登录后查看C币余额
- 我的收藏
- 我的下载
- 下载帮助
最新资源
- 三相交错LLC谐振仿真闭环,Y型联接(图1主回路图),自均流(图2三相谐振电流波形),软开关(图3是原边mos的驱动和DS和电流
- 二阶锥松弛化的无功优化代码 考虑储能系统ESS,有载调压分接头OLTC,离散电容器CB和静止无功补偿SVC设备约束 以购电和网损
- 三相LCL型并网逆变器 MATLAB 内含:SPWM模块,LCL滤波结构,有源阻尼电容电流比例反馈模块,PI控制器模型 采用dq
- 基于分布式驱动电动汽车的路面附着系数估计,分别采用无迹卡尔曼 滤波(UKF)和容积卡尔曼滤波(CKF)对电动汽车四个车轮的路面附
- Simulink仿真:储能锂电池组SOC估算及均衡控制策略研究 关键词:锂电池;不一致性;模糊控制理论;DC DC双向主动均衡
- MATLAB滚动轴承故障诊断程序:采用西楚凯斯大学数据,首先通过变分模态分解(VMD)算法处理,而后分别通过包络谱分析实现故障诊
- 改进A星算法 剔除冗余节点,光滑转折点 对比优化前后路径
- android系统android系统2
- pid-pidpid-pid
- xs-box-client 内网穿透软件-jdk1.8下载与安装教程64
资源上传下载、课程学习等过程中有任何疑问或建议,欢迎提出宝贵意见哦~我们会及时处理!
点击此处反馈
安全验证
文档复制为VIP权益,开通VIP直接复制
信息提交成功