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.
The theme for this year's International Robot Exhibition (IREX) in Tokyo was "Making a Future with Robot." We're not exactly sure what that means, but we're definitely in favor of it, and here are some of the coolest things that we saw. There's one caveat with our IREX coverage, and that's the fact that there was a bit of a language barrier going on most of the time. With the exception of some big international robotics companies, there simply wasn't a lot of information available on many of the robots that we saw. We're following up as best we can, but in the meantime, enjoy this highlight video and gallery that we've put together for you.
Progress in general HRI metrics, while significant, is often in the form of post-experiment questionnaires or expert video analysis. This is a significant hurdle for any intelligent system that aspires to interact with its social environment. In the case of social navigation, robots must react to a dynamic environment. Socially aware robot behavior requires real time quantitative metrics of human subjective experience. This is a vast topic, which we approach by trying to measure how predictable robot actions are from its impact on the paths taken by passers-by. We chose predictability as a first metric because it can be modelled in terms of efficiency, and it affects safety. Thus, it serves as an interesting proof of concept.