Stop-N-Go: Search-based Conflict Resolution for Motion Planning of Multiple Robotic Manipulators

TL;DR

The paper tackles multi-robot manipulation in crowded environments where temporal coordination is essential. It introduces Stop-N-Go, an A*-based temporal adjustment framework that inserts pauses into independently planned trajectories to minimize makespan without altering spatial paths. The approach is shown to be probabilistically complete and, for the basic version, optimal given a set of initial trajectories, with experiments demonstrating superior success rates in dense scenarios compared to baselines. This method enables scalable, temporally coordinated motion planning for teams of high-DOF manipulators in shared workspaces.

Abstract

We address the motion planning problem for multiple robotic manipulators in packed environments where shared workspace can result in goal positions occupied or blocked by other robots unless those other robots move away to make the goal positions free. While planning in a coupled configuration space (C-space) is straightforward, it struggles to scale with the number of robots and often fails to find solutions. Decoupled planning is faster but frequently leads to conflicts between trajectories. We propose a conflict resolution approach that inserts pauses into individually planned trajectories using an A* search strategy to minimize the makespan--the total time until all robots complete their tasks. This method allows some robots to stop, enabling others to move without collisions, and maintains short distances in the C-space. It also effectively handles cases where goal positions are initially blocked by other robots. Experimental results show that our method successfully solves challenging instances where baseline methods fail to find feasible solutions.

Figures (8)

• Figure 1: Example tasks where robots are assigned specific task locations and need to return after completing their tasks. (Left) Robots A and B have overlapping goal positions to pick. (Right) Robots need to follow predefined trajectories (for welding or painting tasks) but their trajectories block each other to move.
• Figure 2: An overview of the proposed method. For each conflict found, two successor nodes are generated to resolve the conflict by inserting pauses. Node A is found to be conflict-free so the search terminates. The nodes shaded with green represent expanded nodes where the one with red bold outline is the solution.
• Figure 3: Interpolation of the initial trajectories for synchronization. Since the C-space of high-DOF robots cannot be visualized, we use the one-dimensional C-space ($q_i$ for $R_i$). (Left) Tthe initial trajectories with non-uniform time intervals. (Right) After interpolation with $t_\text{intv} = 0.1$.
• Figure 4: An illustration of how a detected conflict is resolved by inserting a pause at $t_s = 4$. The conflict is resolved by shifting the trajectory and adding a pause, which allows $\mathcal{T}_2$ to proceed without conflict.
• Figure 5: An example of SearchJumpStep: After resolving a conflict at $t_s = 4$, a binary search checks $q_1^4$ at $t_s + 1$ with $q_2^{10}$ at $t = 10$, finding no conflict. Next, it checks $q_2^7$ at $t = 7$, finding a conflict. The final check at $t = 8$ finds no conflict, setting $t_c$. A pause is then inserted at $q_1^3$ from $t_s = 4$ to $t_c = 7$.