Pushing Everything Everywhere All At Once: Probabilistic Prehensile Pushing

2025-03-18

IEEE Robotics and Automation Letters

AI Summary

This work addresses the problem of *prehensile pushing*—dexterous manipulation under grasp constraints leveraging environmental reaction forces. We propose a probabilistic modeling framework that represents environmental contact points as a discrete probability distribution; entropy minimization drives multi-point contact to automatically converge to a single optimal push point, eliminating the need for computationally expensive mixed-integer programming. The method integrates nonlinear trajectory optimization, entropy regularization, and probabilistic constraint relaxation. It is validated in simulation and on a Franka Panda robotic platform. Compared to baseline approaches, our method achieves an 8× speedup in trajectory computation and reduces optimization cost by 20×, while ensuring stable, physically feasible manipulation across diverse objects. The core contribution is the first formulation of a “full-point concurrent, single-point focused” thrust optimization paradigm—uniquely balancing computational efficiency with physical realizability.

Technology Category

• Intelligent Robots: Manipulation
• Search and Optimization: Sampling/Simulation-based Search
• Planning, Routing, and Scheduling: Mixed Discrete/Continuous Planning

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web

Abstract

We address prehensile pushing, the problem of manipulating a grasped object by pushing against the environment. Our solution is an efficient nonlinear trajectory optimization problem relaxed from an exact mixed integer non-linear trajectory optimization formulation. The critical insight is recasting the external pushers (environment) as a discrete probability distribution instead of binary variables and minimizing the entropy of the distribution. The probabilistic reformulation allows all pushers to be used simultaneously, but at the optimum, the probability mass concentrates onto one due to the entropy minimization. We numerically compare our method against a state-of-the-art sampling-based baseline on a prehensile pushing task. The results demonstrate that our method finds trajectories 8 times faster and at a 20 times lower cost than the baseline. Finally, we demonstrate that a simulated and real Franka Panda robot can successfully manipulate different objects following the trajectories proposed by our method. Supplementary materials are available at https://probabilistic-prehensile-pushing.github.io/.

Problem

• Efficient manipulation of grasped objects via pushing against environment
• Reformulating pushers as discrete probability distributions for optimization
• Achieving faster and lower-cost trajectories for robotic object manipulation

Innovation

• Probabilistic reformulation of external pushers
• Efficient nonlinear trajectory optimization
• Entropy minimization for probability distribution