An Integrated Dynamic Method for Allocating Roles and Planning Tasks for Mixed Human-Robot Teams

Abstract

This paper proposes a novel integrated dynamic method based on Behavior Trees for planning and allocating tasks in mixed human robot teams, suitable for manufacturing environments. The Behavior Tree formulation allows encoding a single job as a compound of different tasks with temporal and logic constraints. In this way, instead of the well-studied offline centralized optimization problem, the role allocation problem is solved with multiple simplified online optimization sub-problem, without complex and cross-schedule task dependencies. These sub-problems are defined as Mixed-Integer Linear Programs, that, according to the worker-actions related costs and the workers' availability, allocate the yet-to-execute tasks among the available workers. To characterize the behavior of the developed method, we opted to perform different simulation experiments in which the results of the action-worker allocation and computational complexity are evaluated. The obtained results, due to the nature of the algorithm and to the possibility of simulating the agents' behavior, should describe well also how the algorithm performs in real experiments.

A Comprehensive Architecture for Dynamic Role Allocation and Collaborative Task Planning in Mixed Human-Robot Teams

“…this algorithm features the custom Behavior Trees presented in subsection II-A as task model and the allocation method that includes also collaborative action execution, explained in subsubsection II-B3. this algorithm exploits both task model and allocation method in [37]. Actions are planned only cooperatively.…”

“…A preliminary version of this architecture was introduced in [37] where BTs were first presented as a method for planning subtasks that could later be allocated by means of different MILPs. However, this version of the method could not deal with i) a general structure of tasks (and hence allowing the maximum degree of action parallelization) and ii) proper human-robot collaborative action executions, where only single agent actions are envisioned (cooperation).…”

“…We compared the time complexity of our approach with the gold standard of role allocation methods for assembly tasks, the And/Or Graph (AOG) [11], [19], to be used as baseline, and the one presented by the authors in a previous work [37], that still exploits BTs as task planners but did not foresee by definition any collaborative task allocation (and hence execution) and does not allow a complete parallelization of tasks. Since also AOGs model assembly tasks where all actions are in sequence, we compared the performances with tasks composed of actions only in sequence.…”

“…To evaluate the planning and allocation capabilities of the new method with respect to the approach in [37] a simulation was performed with a target job composed of 13 actions performed by 3 workers. We assume that each agent executes the assigned task within the specified execution time.…”

Design of Advanced Human–Robot Collaborative Cells for Personalized Human–Robot Collaborations

“…Regarding HRC, human factors have been considered at different levels [3]. For example, optimization of human factors have been embedded into a task scheduler [4]; task allocation have been exploited to reduce the workload of human workers [5,6]; task synergy between human-robot tasks were optimized to reduce the cycle time [7]; human-aware motion planners demonstrated to be preferable by human users [8]. The works mentioned above mainly focus on the planning aspects of HRC.…”