Ergonomic and Worker-Centric Human-Robot Collaboration: Strategies, Interfaces and Controllers

An emergent trend of flexible production lines is represented by the attempt of pairing human workers with cobots. Robotic agents offer high-precision motions and considerable power capacity, whereas human workers can complement the robots with their superior cognitive capabilities and task understanding. The greatest advantage of this match is achievable not only allowing cobots to work side-by-side with humans coexistence but envisioning fully cooperative and, if needed, even collaborative scenarios. Several studies showed that a safe physical coexistence with cobot is not only feasible but could also significantly improve the production process. The researchers' goal in this field consists in providing the cobot with intelligent algorithms and interfaces that allow a fruitful collaboration and physical interaction. The development of a collaborative solution can be split into different levels: at the task level, the specific production process is analysed and decomposed into a sequence of atomic actions. The decomposition is independent of the agents that compose the mixed team. The nature of the team affects the team level, whose strategies tackle problems like role allocation, that defines which agent is in charge of each action, and robotic actions planning and scheduling. At the agent level, we need to ensure, together with robot motion control strategy, agents coordination and intuitive interactions. This thesis aims to face these open problems, focusing in the strategies, interfaces and controllers for Human-Robot Collaboration (HRC) in industrial environments, such as manufacturing and logistics.