Prev-Talk | Next-Talk | All-Talks | Talks-Sorted

Talk-15: Aquaticus - A Human-Robot Capture-the-Flag Competition for Research in Human-Robot Teaming

Michael Novitzky, Michael Benjamin, Paul Robinette, Henrik Schmidt - MECHE/CSAIL, MIT.

Aquaticus is a project with the objective of developing a competition where humans and robots compete together against similarly configured teams of humans and robots. The goal of this project is to better understand the role of human trust in collaborations between humans and robots, how trust is achieved and affects overall team performance, and whether humans explicitly design and configure their robotic teammates with trust in mind. It is our assertion that to truly study the impact of trust on team effectiveness it must be performed in reasonably complex win-or-lose environments. By embedding the competition on the water we introduce stress on the team that will expose interactions that are currently not seen in other human-robot teaming scenarios. Furthermore the environment needs to be low-cost, readily accessible with the ability to allow humans to repeatedly experiment and innovate within the autonomy/trust design space, to learn what is effective and why.

The competition is conducted on the water in about a one square kilometer area off the MIT Sailing Pavilion. Competitors are responsible for designing and building the autonomy algorithms driving their robots and competing alongside their robots and human teammates. This research aims to explore a relatively poorly understood intersection of two well known areas of robotic systems research: the (embedded) human-robot interface and multi-robot zero-sum competitions. The intersection is created by embedding humans alongside robots, as equal teammates, on a playing field competing against a similarly configured team. Autonomy algorithm design and human robot communications will be re-considered from the perspective of the situated human who will be dealing with stress, uncertainty and pressures of performing their own role in the competition, in coordination with their robot teammates.

Categories:

  • ASVs
  • Human-Robot Teaming