Many people may benefit from assistive robots that understand their users' intentions and aid them with the execution of these intentions in a safe and intuitive way through shared control. In the past, our research group has worked on semi-autonomous robotic wheelchairs transporting people with mobility challenges. Experimental results with our user-adaptive Bayesian approach for both intention estimation and shared human-machine decision-making under uncertainty have shown that in situations where the driver changes his or her intention, the assistive behavior by the robot may under certain conditions be counter-intuitive as it continues to take actions that are in line with the previous user intention, and this for too long a period of time. To remedy this, this paper proposes an approach to detect such changes in user plans in order to make the robot's assistive behavior more reactive and thus more intuitive. The approach adopts a test that checks the consistency of the posterior distribution over user intentions with the given steering signals.

Developing collaborative robots that can productively operate out of isolation and work safely in uninstrumented, human-populated environments is critically important for advancing the field of robotics. Especially in domains where modern robots are ineffective, we wish to leverage human-robot teaming to improve the efficiency, ability, and safety of human workers. Our work, outlined in this extended abstract, focuses on creating agents capable of human-robot teamwork by leveraging learning from demonstration, hierarchical task networks, multi-agent planning and state estimation, and intention recognition. We briefly describe our recent work within human-robot collaboration, including task comprehension, learning and performing assistive behaviors, and training novice human collaborators to become competent co-workers.