NRI: Collaborative Research: Co-Exploration using Science Hypothesis Maps

Carnegie Mellon University, Pittsburgh PA

Investigators

Abstract

This project studies a new approach to exploration. In this approach, the explorer's description of where to go is not prescribed by a path, but instead by a model of what the explorer believes. In the context of a human scientist and robotic explorer working together, this approach transforms the co-robotic relationship into a collaboration in which the human and robot work together to fill in gaps in knowledge to make discoveries. The science hypothesis map is the structure in which human scientists initially describe their belief about the world and in which the belief state evolves as the robotic explorer collects information. This project addresses a pressing and understudied area of human/robot interaction: it directly tackles the challenges of human scientist interaction with remote robotic explorer by crafting a representation of the environment that is expressive and intuitive for the human while quantifying information to enable meaningful robotic decisions. The hypothesis map formulation is a novel and radical departure from our prior work which has focused on human goal selection and robotic autonomy. It has the potential to revolutionize exploration robotics by changing the nature of the relationship between human and robot from supervisory control to shared information discovery.