A crucial aspect of physical human–robot collaboration (HRC) is to maintain a safe common workspace for human operator. However, close proximity between human–robot and unpredictability of human behavior raises serious challenges in terms of safety. This article proposes a risk analysis methodology for collaborative robotic applications, which is compatible with well-known standards in the area and relies on formal verification techniques to automate the traditional risk analysis methods. In particular, the methodology relies on temporal logic-based models to describe the different possible ways in which tasks can be carried out, and on fully automated formal verification techniques to explore the corresponding state space to detect and modify the hazardous situations at early stages of system design.

中文翻译：

---

通过自动形式验证对协作机器人进行安全评估

物理人机协作 (HRC) 的一个关键方面是为人类操作员维护一个安全的公共工作空间。然而，人与机器人之间的密切接触以及人类行为的不可预测性在安全方面提出了严峻的挑战。本文提出了一种协作机器人应用的风险分析方法，该方法与该领域的知名标准兼容，并依靠形式验证技术使传统的风险分析方法自动化。特别是，该方法依赖于基于时间逻辑的模型来描述执行任务的不同可能方式，并依赖于全自动形式验证技术来探索相应的状态空间以检测和修改早期阶段的危险情况。系统设计。