Adapting Human Leadership Approaches for Role Allocation in Human-Robot Navigation Scenarios

Abstract

In this paper, we propose to examine the practices of leadership defined in human relationships and model their use in maximizing performance for human-robot interaction scenarios. This process involves first defining the human-robot space of interaction and mapping the situational context in which human leadership styles are most fitting. We then determine which behavior, for both the human and robot, is most appropriate in order to understand the proper roles for human-robot integration. From there, we model the necessary robot behavior for increasing efficiency in human-robot interaction schemes. We conclude by discussing experimental results derived from allocating roles in representative human-robot navigation scenarios.

A meta-analysis of human-system interfaces in unmanned aerial vehicle (UAV) swarm management

“…By balancing between the cognitive limits of the operator and the swarm's need for guidance and specific decisions, the researchers were able to control swarms of up to 100 UAVs without an increase in errors or loss of SA. Howard and Cruz (2006) had similar findings in a related experiment where they tested directive (or command) leadership's effectiveness against that of transactional (or reward-based, collaborative) leadership. They also found that operators functioned best when guiding the swarm as part of the team, rather than as a commander, in part because it allowed the operator to focus on a manageable section of the swarm, rather than the entire group.…”

“…By using transactional control (incentivizing the swarm into performing a task without commanding each step of taskcompletion process) on swarms with a higher LOA, Howard and Cruz (2006) showed that swarms could achieve their goal with fewer errors and lower cognitive workload for the operator. Kira and Potter (2009) corroborated these findings in a scenario requiring the swarm to split up while defending a point and pursuing an attacker.…”

Human leadership approaches to hybrid human/robot teams

“…The theory states that humans adjust to technology and, similarly, technology should adapt to humans. Following this theory and the work of [8], we can justify an attempt to apply leadership styles developed for human organisations to the human-robot interaction space. It is hoped that the wellunderstood nature of these leadership styles and the familiarity of their application will resonate with human operators and lead to a greater level of comfort and acceptance in hybrid human/robot teams.…”

“…Perhaps the complement of this application of computer science to humans is [8] where human leadership paradigms are examined for relevance to the case where the team's leader is leading a single robot. In this work, the authors collect common statements of human leadership styles from [12] [14][3] and attempt to remove reference to human traits such as personality, reducing the style to a set of concepts which can be implemented in a robot control interface.…”

“…In this paper the concepts in [8] are extended to a common application of human/robot teams -the exploration and navigation of an area. The roles of human leader and robot team members are defined across a range of common leadership styles and the situations in which each leadership style could be applied are explored.…”