The Fourth Industrial Revolution will lead to a step change in the level of autonomy in manufacturing facilities. Autonomy involves enabling machines to make decisions that were previously made by humans. This movement of responsibility for decision-making from humans to machines raises a number of challenges when assuring the safety of the operations. The challenges are exacerbated further by the use of artificial intelligence in system training and control, and by the greater use of robots in areas where there is close interaction between humans and machines (often referred to as cobots).
In this paper, we first provide a general discussion of the challenges of autonomy, and then we use three scenarios for manufacturing, representing a progression through different levels of automation and autonomy. This highlights how the nature and extent of the safety assurance challenge increase for each scenario. We discuss possible approaches for addressing the issues and identify the key open research challenges.