Cooperative Task and Motion Planning for Multi-Arm Assembly Systems

Chen, Jingkai; Li, Jiaoyang; Huang, Yijiang; Garrett, Caelan Reed; Sun, Dalin; Fan, Chuchu; Hofmann, Andreas; Mueller, Caitlin; Koenig, Sven; Williams, Brian

doi:10.48550/arxiv.2203.02475

Cited by 4 publications

(6 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We did not put an emphasis on comparison with other methods, as the other methods we are aware of ( [8], [11], [12], [25], [58], [68]) do not scale to the number of robots and objects we consider. We compare our method to a fixed-time sampling scheme, 5 which decomposes the problem as the presented approach does, and uses prioritized planning, but does not allow for variable durations of the tasks, i.e., all mode-switches take place at a multiple of T .…”

Section: Comparison To Fixed-time-samplingmentioning

confidence: 99%

Long-Horizon Multi-Robot Rearrangement Planning for Construction Assembly

et al. 2023

View full text Add to dashboard Cite

Robotic construction assembly planning aims to find feasible assembly sequences as well as the corresponding robotpaths and can be seen as a special case of task and motion planning (TAMP). As construction assembly can well be parallelized, it is desirable to plan for multiple robots acting concurrently. Solving TAMP instances with many robots and over a long time-horizon is challenging due to coordination constraints, and the difficulty of choosing the right task assignment. We present a planning system which enables parallelization of complex task and motion planning problems by iteratively solving smaller subproblems. Combining optimization methods to jointly solve for manipulation constraints with a sampling-based bi-directional space-time path planner enables us to plan cooperative multi-robot manipulation with unknown arrival-times. Thus, our solver allows for completing subproblems and tasks with differing timescales and synchronizes them effectively. We demonstrate the approach on multiple construction case-studies to show the robustness over long planning horizons and scalability to many objects and agents. Finally, we also demonstrate the execution of the computed plans on two robot arms to showcase the feasibility in the real world.

show abstract

Section: Comparison To Fixed-time-samplingmentioning

confidence: 99%

Long-Horizon Multi-Robot Rearrangement Planning for Construction Assembly

et al. 2023

View full text Add to dashboard Cite

show abstract

“…A further extension is to simultaneously plan motions for multiple robots as in [19] and [20]. In [21] also the fixed sequences are challenged by considering some robot-robot collisions during scheduling, the remaining robot-robot collisions are prevented by using multi-robot motion planning in a post processing step.…”

Section: Related Workmentioning

confidence: 99%

Spatial–Temporal Load Balancing and Coordination of Multi-Robot Stations

Åblad

Spensieri

Bohlin

et al. 2023

IEEE Trans. Automat. Sci. Eng.

View full text Add to dashboard Cite

Cycle time minimization in multi-robot manufacturing stations is computationally challenging. This is due to the many aspects that need to be accounted for, including assigning process tasks to robots, specifying robot configurations at tasks, sequencing, planning motions, and coordinating the robots to avoid collisions. Hence, to find good solutions, often some assumptions are made and/or the problem is divided into subproblems-often limiting the set of solutions with the risk of excluding the best ones. In this study, we generalize the completely disjoint solution method that challenges the socalled shortest path assumption, i.e., to let each robot use its shortest collision-free motion between any two configurations, regardless of the other robots. We devise a generalized method called spatial-temporal load balancing and coordination, which prevents robot-robot collisions by a sequence of disjoint solutions, guiding task assignments, sequences, and robot motions (path and velocity). We study both artificial and industrial instances. For some of them, our suggested method is superior to methods based on the shortest path assumption, with as much as a 28% reduction in cycle time. Moreover, for problem instances with no special structure, we establish that the shortest path assumption is often reasonable. Note toPractitioners-This work is motivated by a particular industrial problem instance of a spot-welding station with two robots and where welds are placed along the edge of a workpiece. Due to the special geometry of the instance one robot can only perform welds in the middle of the edge and the other only at the ends. As a result, if the robots use their shortest motions between welds, then waiting times are required to prevent collisions. Moreover, the tasks are too close to each other to allow for a completely disjoint solution. Hence, we suggest a method based on sequence of disjoint solutions.

show abstract

“…In [18], [19], efficient approaches are proposed for the multi-robot object retrieval problem, assuming permanent object removal and considering one target object at a time, while our planner considers several target objects at the same time and relocates the obstacles within the workspace. Multi-robot rearrangement planning problems [5]- [7] are also closely related to MR-GTAMP. However, the rearrangement planning problems assume that the goal configurations of the objects are given, while MR-GTAMP requires the planners to decide which objects to move and to which positions.…”

Section: Related Workmentioning

confidence: 99%

“…MR-GTAMP naturally arises in many multi-robot manipulation domains, such as multi-robot construction, assembly and autonomous warehousing [5], [6]. MR-GTAMP is interesting as multi-robot systems can perform manipulation tasks more effectively than single-robot systems and can also perform manipulation tasks that are beyond the capabilities of single-robot systems [7].…”

Section: Introductionmentioning

confidence: 99%

“…Then, for each pair of a robot R ∈ R and its grasp g M,R ∈ Gr M,R , we find all partially grounded pick-and-place actions ā that move object M to its target region Re M with R as the pick robot (Alg. 1, line [4][5][6][7][8][9][10][11][12][13][14][15][16][17][18]. For each partially grounded pick-and-place action ā, we find all movable objects that block the pick action of ā and add the corresponding block-pick edges (Alg.…”

mentioning

confidence: 99%

See 1 more Smart Citation

A MIP-Based Approach for Multi-Robot Geometric Task-and-Motion Planning

Zhang¹,

Chan²,

Zhong³

et al. 2022

Preprint

View full text Add to dashboard Cite

We address multi-robot geometric task-andmotion planning (MR-GTAMP) problems in synchronous, monotone setups. The goal of the MR-GTAMP problem is to move objects with multiple robots to goal regions in the presence of other movable objects. To perform the tasks successfully and effectively, the robots have to adopt intelligent collaboration strategies, i.e., decide which robot should move which objects to which positions, and perform collaborative actions, such as handovers. To endow robots with these collaboration capabilities, we propose to first collect occlusion and reachability information for each robot as well as information about whether two robots can perform a handover action by calling motionplanning algorithms. We then propose a method that uses the collected information to build a graph structure which captures the precedence of the manipulations of different objects and supports the implementation of a mixed-integer program to guide the search for highly effective collaborative task-andmotion plans. The search process for collaborative task-andmotion plans is based on a Monte-Carlo Tree Search (MCTS) exploration strategy to achieve exploration-exploitation balance. We evaluate our framework in two challenging GTAMP domains and show that it can generate high-quality task-andmotion plans with respect to the planning time, the resulting plan length and the number of objects moved compared to two state-of-the-art baselines.

show abstract

Cooperative Task and Motion Planning for Multi-Arm Assembly Systems

Cited by 4 publications

References 20 publications

Long-Horizon Multi-Robot Rearrangement Planning for Construction Assembly

Long-Horizon Multi-Robot Rearrangement Planning for Construction Assembly

Spatial–Temporal Load Balancing and Coordination of Multi-Robot Stations

A MIP-Based Approach for Multi-Robot Geometric Task-and-Motion Planning

Contact Info

Product

Resources

About