Abstract
We consider the nonprehensile object transportation task known as the waiter's problem - in which a robot must move an object on a tray from one location to another - when the transported object has uncertain inertial parameters. In contrast to existing approaches that completely ignore uncertainty in the inertia matrix or which only consider small parameter errors, we are interested in pushing the limits of the amount of inertial parameter uncertainty that can be handled. We first show how constraints that are robust to inertial parameter uncertainty can be incorporated into an optimization-based motion planning framework to transport objects while moving quickly. Next, we develop necessary conditions for the inertial parameters to be realizable on a bounding shape based on moment relaxations, allowing us to verify whether a trajectory will violate the constraints for any realizable inertial parameters. Finally, we demonstrate our approach on a mobile manipulator in simulations and real hardware experiments: our proposed robust constraints consistently successfully transport a 56 cm tall object with substantial inertial parameter uncertainty in the real world, while the baseline approaches drop the object while transporting it.
Introduction
In robotics, "nonprehensile" manipulation refers to moving objects without a firm grasp, such as pushing, sliding, or balancing. A classic example is the waiter’s problem, where a robot must move a tray quickly without the items sliding or tipping. The primary challenge lies in the uncertainty of the object's internal mass distribution. While a robot can perceive an object's outer dimensions, it often lacks knowledge of the internal center of mass (CoM) and inertia matrix. If a box is top-heavy or has an offset CoM, standard motion plans may accelerate too aggressively, leading to failure.
Methodology
This paper introduces a two-part framework to address these uncertainties:
- Robust Motion Planning: The task is formulated as a Constrained Optimal Control Problem (OCP). The researchers define a polyhedral set of possible CoM locations and enforce "sticking constraints"—mathematical conditions ensuring the object doesn't slide or tip—for every vertex of that set. By satisfying these constraints at the extremes, the robot ensures stability for any CoM within the defined volume.
- Verification via Moment Relaxations: To account for uncertainty in the inertia matrix, the authors utilize moment relaxations (Lasserre’s hierarchy). They developed a Semidefinite Program (SDP) to verify if a planned trajectory remains safe for any physically realizable mass distribution within the object's shape, providing a theoretical guarantee against dropping the object.
Experimental Results
The framework was evaluated through extensive simulations and hardware experiments. In 1,620 PyBullet simulation runs using a 9-DOF mobile manipulator, the robust planner achieved a 100% success rate across all tested object heights (30 cm to 60 cm).
As illustrated in the results, baseline approaches that assume the CoM is at the geometric center or top of the object experience significant performance degradation as height increases. In real-world hardware tests, the robust planner successfully transported a 56 cm tall box with a hidden, offset CoM (created using sugar-filled bottles), whereas baseline methods dropped the box almost immediately.
Conclusion and Future Work
The proposed robust constraints allow for dynamic transportation of objects even with substantial inertial uncertainty. While the current trajectories are calculated offline, future research will focus on improving computational efficiency—specifically for the SDP verification—to enable real-time reactive planning in dynamic environments.
Summary updated