Evaluating path planning in human-robot teams: Quantifying path agreement and mental model congruency

Abstract

The integration of robotic systems into daily life is increasing, as technological advancements facilitate independent and interdependent decision-making by autonomous agents. Highly collaborative human-robot teams promise to maximize the capabilities of humans and machines. While a great deal of progress has been made toward developing efficient spatial path planning algorithms for robots, comparatively less attention has been paid to developing reliable means by which to assess the similarities and differences in path planning decisions and associated behaviors of humans and robots in these teams. This paper discusses a tool, the Algorithm for finding the Least Cost Areal Mapping between Paths (ALCAMP), which can be used to compare paths planned by humans and algorithms in order to quantify the differences between them, and understand the user's mental models underlying those decisions. In addition, this paper discusses prior and proposed future research related to human-robot collaborative teams. Prior studies using ALCAMP have measured path divergence in order to quantify error, infer decision-making processes, assess path memory, and assess team communication performance. Future research related to human-robot teaming includes measuring formation and path adherence, testing the repeatability of navigation algorithms and the clarity of communicated navigation instructions, inferring shared mental models for navigation among members of a group, and detecting anomalous movement.