Coordinated Multiagent Teams and Sliding Autonomy for Large-Scale Assembly

Abstract

Recent research in human-robot interaction has investigated the idea of Sliding, or Adjustable, Autonomy, a mode of operation bridging the gap between complete robot autonomy and full teleoperation.This work, by and large, has been in single-agent domains -involving only one human and one robot -and has not examined the issues that arise when moving to multi-agent domains. Here, we discuss the issues involved when adapting Sliding Autonomy concepts to coordinated multi-agent teams. In our system, remote human operators have the ability to join, or leave, the team at will, to assist the autonomous agents with their tasks while not disrupting the team's coordination. We employ user modeling in order to allow agents to request help when appropriate, regardless of whether human operators are actively monitoring their progress. To validate our approach, we present the results of two experiments. The first evaluates the human-multi-robot team's performance under four different collaboration strategies including complete teleoperation, pure autonomy, and two distinct versions of Sliding Autonomy. The second experiment compares a variety of user interface configurations, to investigate how quickly a human operator can attain situational awareness when asked to help. The results of these studies support our belief that by incorporating a remote human operator into multi-agent teams, the team as a whole becomes more robust and efficient.

Autonomous construction using scarce resources in unknown environments

“…Some of these target construction systems for specific tasks, such as structure assembly in extraterrestrial settings (Stroupe et al 2006;Huntsberger et al 2003), in orbit (Everist et al 2004;Ueno et al 2003), or for assembling vertical trusses (Lindsey et al 2011). Some works explore how to integrate a human in the loop to handle contingencies (Sellner et al 2006). With the same engineering orientation, architects (Gramazio 2008) and the construction industry (Gambao et al 2000) have shown a long-lasting interest in automation and robotic approaches (Skibniewski and Wooldridge 1992;Zavadskas 2010).…”

Improving Task Performance through High Level Shared Control of Multiple Robots with a Context Aware Human-Robot Interface

“…The literature for traded control has typically focused on trading control between full autonomous control and full manual (user) control [12], [13]. In [12], the autonomous agent would take control when precise movements are needed during grasping with the user in full control otherwise.…”

“…In [12], the autonomous agent would take control when precise movements are needed during grasping with the user in full control otherwise. Examining whether the user or the autonomous agent should initiate traded control was examined in [13]. The user could take control if they determined that the robot was struggling with a task.…”

“…The allocation authority can be part of the responsibilities of the autonomous agent or the user, or it might be an independent function that is separate from either. Allocation authority has been investigated in low level traded control where either the autonomous agent or user acted as the allocation authority [13]. They found that an approach where either the human operator can interrupt the system to take control or the system can request help if needed had a higher task completion time and success rate than an approach where only the system can request help.…”

Intelligent Multiagent Coordination Based On Neuro-Fuzzy Models With Hierarchical Reinforcement Learning

“…A good coordination system should avoid or minimize the occurrence of these problems by optimizing the use of resources and the time required to achieve the objectives [14]. Effective coordination between agents operating in a shared environment contributes to improving the quality of the solutions achieved and to improved performance in solving tasks [15].…”