Integrated Topological Planning and Scheduling
25
between agents, and between humans and robots in specific. In this context, this
paper introduces an HRC orchestration problem definition that considers collaboration features at a large scale. Furthermore, the proposed model considers
topological placement of resources in a novel approach that respects factory floor
constraints. In addition, we derive an abstract model for accurate approximate
evaluation of topological and task assignment scenarios that is simple and fast
and as such applicable to stochastic optimization scenarios. Finally, we present
the application of the above in a comprehensive optimization workflow for deriving detailed agent assignment and scheduling solutions for large scale HRC.
We emphasize that our work focuses on large-scale distributed collaboration,
where it is possible to distribute agents to more than a single workstation as
required to improve scheduling. In an HRC setting, for instance, such a scenario
may involve robot manipulators in each workstation of a factory floor that perform the bulk of the work, while humans are distributed overseeing workstation
clusters, supervising each workstation and performing few tasks that cannot be
performed by robots. To the best of the authors’ knowledge, there is no other
work that addresses this requirement specifically.
This new role of human workers, which leans more towards the supervisory
side, has the potential to shift labor conditions in industry, by enabling skilled
workers to make use of their skills rather than labor, while at the same time
assuming a more responsible role in the production line.
1.2 Paper Outline
The paper is structured as follows: In Sect. 2 a concise review of related work
in orchestration of large-scale HRC teams is presented. Section 3 elaborates on
the proposed approach, presenting a formal problem definition and a detailed
outline of the proposed optimization strategy to address the problem. Section 4
presents a series of preliminary validation experiments. Section 5 provides an
overall discussion on derived results. Finally, Sect. 6 concludes the paper.
2 Related Work
2.1 Orchestration of Large HR Teams
Orchestration or robot teams in industry is a distinct research topic that has
received significant attention in literature for a long time. This topic considers
the scheduling of teams comprising multiple robots to efficiently perform tasks
that may be carried out in parallel or that may involve collaboration between
individual robots. Orchestration of human-robot teams, on the contrary, has
received less attention so far.
In [7], authors propose a multi-stage optimization process for task assignment
and scheduling of a multiple robot team factory floor. They propose a series of
problem abstractions that are solved with each step incorporating more problem
details than the last. Authors report improved results compared to heuristic and
pure MILP solutions.
25
between agents, and between humans and robots in specific. In this context, this
paper introduces an HRC orchestration problem definition that considers collaboration features at a large scale. Furthermore, the proposed model considers
topological placement of resources in a novel approach that respects factory floor
constraints. In addition, we derive an abstract model for accurate approximate
evaluation of topological and task assignment scenarios that is simple and fast
and as such applicable to stochastic optimization scenarios. Finally, we present
the application of the above in a comprehensive optimization workflow for deriving detailed agent assignment and scheduling solutions for large scale HRC.
We emphasize that our work focuses on large-scale distributed collaboration,
where it is possible to distribute agents to more than a single workstation as
required to improve scheduling. In an HRC setting, for instance, such a scenario
may involve robot manipulators in each workstation of a factory floor that perform the bulk of the work, while humans are distributed overseeing workstation
clusters, supervising each workstation and performing few tasks that cannot be
performed by robots. To the best of the authors’ knowledge, there is no other
work that addresses this requirement specifically.
This new role of human workers, which leans more towards the supervisory
side, has the potential to shift labor conditions in industry, by enabling skilled
workers to make use of their skills rather than labor, while at the same time
assuming a more responsible role in the production line.
1.2 Paper Outline
The paper is structured as follows: In Sect. 2 a concise review of related work
in orchestration of large-scale HRC teams is presented. Section 3 elaborates on
the proposed approach, presenting a formal problem definition and a detailed
outline of the proposed optimization strategy to address the problem. Section 4
presents a series of preliminary validation experiments. Section 5 provides an
overall discussion on derived results. Finally, Sect. 6 concludes the paper.
2 Related Work
2.1 Orchestration of Large HR Teams
Orchestration or robot teams in industry is a distinct research topic that has
received significant attention in literature for a long time. This topic considers
the scheduling of teams comprising multiple robots to efficiently perform tasks
that may be carried out in parallel or that may involve collaboration between
individual robots. Orchestration of human-robot teams, on the contrary, has
received less attention so far.
In [7], authors propose a multi-stage optimization process for task assignment
and scheduling of a multiple robot team factory floor. They propose a series of
problem abstractions that are solved with each step incorporating more problem
details than the last. Authors report improved results compared to heuristic and
pure MILP solutions.