Collaborative human-computer decision support for planetary surface traversal
In support of the recent vision for humans to return to the Moon, we are studying decision support for planetary surface traversals. A key aspect of this investigation is the appropriate role allocations and balance between automation and human participation in the decision process. Some degree of automation is necessary during on-route replanning because there are many variables and models to consider within a short span of time. Yet, keeping the human in the decision-making loop is critical because astronauts will be able to more readily adapt to change and provide flexibility in problem solving tasks under unexpected circumstances. In this paper, we present a prototype path planner that allows human-automation interaction in the attempt to plan and optimize paths based on objective functions important to planetary traversals. One promising path planning visualization technique is based on the numerical potential field method (NPFM), which communicates to the user how the automation calculates least-costly paths. Our pilot experiment demonstrated that even with simple cost predictions for planned paths, all subjects planned a route within 25% of the optimal route. The results also suggest the NPFM visualization was particularly helpful for subjects tasked to create least-costly paths for a complex objective function.
