Safety assurance at GM and using model management to support it

在设计汽车时，汽车制造商--原始设备制造商(OEM)--会尽一切努力 以确保车辆的安全可靠。随着车辆复杂性的增加， 软件驱动的混合动力系统和高级驾驶员辅助系统的系统设计和安全性 各进程所需的工作量和难度也相应增加。最近，看似次要的 车辆软件和/或物理设计的增量更改已导致安全召回，原因是 不可预见的子系统相互作用已将车辆的安全性和/或可靠性降低到不可接受的水平。至 为了解决这些问题，OEM及其供应商合作生产了标准的ISO 26262路 车辆-功能安全，它规定了应作为以下内容生产的过程及其输出 要求不同汽车安全完整性的车辆功能的系统、软件和安全流程 级别。随着OEM和供应商努力与ISO 26262保持一致，一个日益重要的问题 产生：什么时候可以安全地重用设计和安全保证制品，什么时候工艺和 由此产生的工作产品必须重做吗？汽车公司必须解决这个关键问题，如果他们想要 保持竞争力，并能够以公众愿意支付的购买成本交付更安全的车辆 付钱。为了帮助加拿大汽车业解决这个问题，该项目将对设计和安全进行建模 人工制品及其与明确的安全保证论点的关系称为(安全)保证案例。 我们将提供模型管理流程和工具来分析模型并确定 保险箱的设计变更。这将帮助工程师确定保证案例的哪些部分可以 安全地重复使用，并且需要进一步的工程工作来证明新系统是 安全得可以接受。这项工作将扩展到管理整个产品线的安全保证模型，以 帮助原始设备制造商应对全球产品开发的制造现实。

When a vehicle is designed, the car maker - the Original Equipment Manufacturer (OEM) - makes every effort to ensure that the vehicle will be safe and reliable. As vehicles have increased in complexity with the addition of software-enabled hybrid powertrains and Advanced Driver Assistance Systems, the system design and safety processes have had a corresponding increase in required effort and difficulty. Recently, seemingly minor incremental changes in the vehicle's software and/or physical design have resulted in safety recalls because of unforeseen subsystem interactions that have reduced vehicle safety and/or reliability to unacceptable levels. To help address these issues, OEMs and their suppliers collaborated to produce the standard ISO 26262 Road Vehicles - Functional Safety, which prescribes processes and their outputs that should be produced as part of the system, software and safety processes for vehicle features that require different Automotive Safety Integrity Levels. As OEMs and suppliers strive to be consistent with ISO 26262, an increasingly important question arises: when can design and safety assurance artefacts be safely reused and when do the processes and the resulting work products have to be redone? Automotive companies must address this key question if they want to remain competitive and be able to deliver ever safer vehicles at a purchase cost that the public is willing to pay. To help the Canadian automotive industry address this issue, this project will model the design and safety artefacts and their relationships with explicit safety assurance arguments known as a (safety) assurance case. We will provide model management processes and tools to analyze the models and determine the impact of a design change on the assurance case. This will help engineers determine which parts of the assurance case can be safely reused and where it requires further engineering effort to demonstrate that the new system is acceptably safe. The work will be extended to manage models of safety assurance for entire product lines to help OEMs deal with the reality of manufacturing for global product development.