Collaborative Planar Pushing of Polytopic Objects with Multiple Robots in Complex Scenes

TL;DR

This work tackles collaborative planar pushing of polytopic objects by multiple robots in cluttered environments. It advances a hybrid planning framework that combines quasi-static feasibility, arc-transition modes, and a hierarchical KG-HS search with online NMPC tracking, together with an adaptation mechanism for unknown disturbances. Theoretical guarantees on completeness and feasibility are provided, and extensive simulations and hardware experiments demonstrate robustness, scalability, and practical efficacy across diverse object shapes and fleet sizes. The proposed approach enables efficient, reliable multi-robot manipulation in realistic scenes and lays groundwork for extending to heterogeneous fleets and 3D pushing scenarios.

Abstract

Pushing is a simple yet effective skill for robots to interact with and further change the environment. Related work has been mostly focused on utilizing it as a non-prehensile manipulation primitive for a robotic manipulator. However, it can also be beneficial for low-cost mobile robots that are not equipped with a manipulator. This work tackles the general problem of controlling a team of mobile robots to push collaboratively polytopic objects within complex obstacle-cluttered environments. It incorporates several characteristic challenges for contact-rich tasks such as the hybrid switching among different contact modes and under-actuation due to constrained contact forces. The proposed method is based on hybrid optimization over a sequence of possible modes and the associated pushing forces, where (i) a set of sufficient modes is generated with a multi-directional feasibility estimation, based on quasi-static analyses for general objects and any number of robots; (ii) a hierarchical hybrid search algorithm is designed to iteratively decompose the navigation path via arc segments and select the optimal parameterized mode; and (iii) a nonlinear model predictive controller is proposed to track the desired pushing velocities adaptively online for each robot. The proposed framework is complete under mild assumptions. Its efficiency and effectiveness are validated in high-fidelity simulations and hardware experiments. Robustness to motion and actuation uncertainties is also demonstrated.

Figures (23)

• Figure 1: Snapshots of the collaborative pushing task in simulation and hardware experiments.
• Figure 2: Overall framework, including the hybrid plan generation in Sec. \ref{['subsec:hybrid']} and the online execution in Sec. \ref{['subsec:control']}.
• Figure 3: Comparison of $\mathrm{J}_\texttt{F}$ in \ref{['eq:feasibility']} and $\mathrm{J}_{\texttt{MF}}$ in \ref{['eq:multi-directional-feasibility']} for different directions between different modes (robots as blue dots and object as a white polygon).
• Figure 4: Arc transitions (Left) and the set of allowed velocities $\mathcal{P}_{\xi}^{{}{\mathrm{B}}}$ (Right) visualized on a unit sphere.
• Figure 5: Illustration of the mode generation process via sparse optimization in \ref{['eq:sparse-opt']}. Left: the optimized matrix $\mathbf{F}^\star_{\mathbf{D}}$ of dimention $N_{\mathcal{V}}\times 2D$, where the magnitude of elements is represented by the color intensity; Right: the best mode selected for the given arc transition within $\mathbf{\Xi}( \overgroup{\mathbf{s}_0\mathbf{s}_{\bar{t}}})$.

Theorems & Definitions (20)

• Remark 1
• Remark 2
• Remark 3
• Remark 4
• Remark 5
• Remark 6
• Lemma 1
• proof
• Lemma 2
• proof