人間中心の動的スケジューリングアーキテクチャによる協調作業の最適化
A Human-Centered Dynamic Scheduling Architecture for Collaborative Application
人間とロボットが協働する作業現場で、作業品質や人間の作業ばらつきを考慮しつつ、リアルタイム監視に基づいてタスクを動的に割り当て・スケジューリングする二層構造のアーキテクチャを提案し、実験で有効性を示した論文。
著者: Andrea Pupa, Wietse Van Dijk, Cristian Secchi
分類: cs.RO
原文アブストラクト
In collaborative robotic applications, human and robot have to work together during a whole shift for executing a sequence of jobs. The performance of the human robot team can be enhanced by scheduling the right tasks to the human and the robot. The scheduling should consider the task execution constraints, the variability in the task execution by the human, and the job quality of the human. Therefore, it is necessary to dynamically schedule the assigned tasks. In this paper, we propose a two-layered architecture for task allocation and scheduling in a collaborative cell. Job quality is explicitly considered during the allocation of the tasks and over a sequence of jobs. The tasks are dynamically scheduled based on the real time monitoring of the human's activities. The effectiveness of the proposed architecture is experimentally validated.
