The challenges of human-robot collaboration arise from the fact that while both agents must act, the robot knows less than the human about what they want. The difficulty lies in making sense of human behavior enough to be able to optimally assist them with autonomy, despite the development of robot capabilities.
The use of Inverse Reinforcement Learning can help robotics achieve more autonomy in assisting people, such as controlling a wheelchair-mounted arm, by learning from demonstrated human behavior.
The challenge of human-robot collaboration lies in understanding human behavior and needs, which is daunting as robot capabilities continue to improve. However, by allowing robots to optimize based on what the human wants, rather than what a programmer dictates, the robot may be able to influence and better understand human behavior.