Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Pushing in the Dark: A Reactive Pushing Strategy for Mobile Robots Using Tactile Feedback

Idil Ozdamar, Doganay Sirintuna, Robin Arbaud, Arash Ajoudani

TL;DR

The paper tackles pushing unknown objects to specific destinations in cluttered environments without relying on object motion models. It introduces the Reactive Pushing Strategy (RPS), a tactile-feedback–driven controller that adaptively modulates base velocities and uses a Realignment State to maintain contact, all under a quasi-static pushing assumption. Across both simulations and real experiments, RPS outperforms non-reactive baselines and a state-of-the-art adaptive pushing method, achieving high success rates even for challenging behind-the-robot targets and varying object properties. The work demonstrates a practical, model-free approach for reliable non-prehensile manipulation with potential for deployment in unstructured settings using only tactile sensing.

Abstract

For mobile robots, navigating cluttered or dynamic environments often necessitates non-prehensile manipulation, particularly when faced with objects that are too large, irregular, or fragile to grasp. The unpredictable behavior and varying physical properties of these objects significantly complicate manipulation tasks. To address this challenge, this manuscript proposes a novel Reactive Pushing Strategy. This strategy allows a mobile robot to dynamically adjust its base movements in real-time to achieve successful pushing maneuvers towards a target location. Notably, our strategy adapts the robot motion based on changes in contact location obtained through the tactile sensor covering the base, avoiding dependence on object-related assumptions and its modeled behavior. The effectiveness of the Reactive Pushing Strategy was initially evaluated in the simulation environment, where it significantly outperformed the compared baseline approaches. Following this, we validated the proposed strategy through real-world experiments, demonstrating the robot capability to push objects to the target points located in the entire vicinity of the robot. In both simulation and real-world experiments, the object-specific properties (shape, mass, friction, inertia) were altered along with the changes in target locations to assess the robustness of the proposed method comprehensively.

Pushing in the Dark: A Reactive Pushing Strategy for Mobile Robots Using Tactile Feedback

TL;DR

The paper tackles pushing unknown objects to specific destinations in cluttered environments without relying on object motion models. It introduces the Reactive Pushing Strategy (RPS), a tactile-feedback–driven controller that adaptively modulates base velocities and uses a Realignment State to maintain contact, all under a quasi-static pushing assumption. Across both simulations and real experiments, RPS outperforms non-reactive baselines and a state-of-the-art adaptive pushing method, achieving high success rates even for challenging behind-the-robot targets and varying object properties. The work demonstrates a practical, model-free approach for reliable non-prehensile manipulation with potential for deployment in unstructured settings using only tactile sensing.

Abstract

For mobile robots, navigating cluttered or dynamic environments often necessitates non-prehensile manipulation, particularly when faced with objects that are too large, irregular, or fragile to grasp. The unpredictable behavior and varying physical properties of these objects significantly complicate manipulation tasks. To address this challenge, this manuscript proposes a novel Reactive Pushing Strategy. This strategy allows a mobile robot to dynamically adjust its base movements in real-time to achieve successful pushing maneuvers towards a target location. Notably, our strategy adapts the robot motion based on changes in contact location obtained through the tactile sensor covering the base, avoiding dependence on object-related assumptions and its modeled behavior. The effectiveness of the Reactive Pushing Strategy was initially evaluated in the simulation environment, where it significantly outperformed the compared baseline approaches. Following this, we validated the proposed strategy through real-world experiments, demonstrating the robot capability to push objects to the target points located in the entire vicinity of the robot. In both simulation and real-world experiments, the object-specific properties (shape, mass, friction, inertia) were altered along with the changes in target locations to assess the robustness of the proposed method comprehensively.
Paper Structure (10 sections, 11 equations, 7 figures, 2 tables, 1 algorithm)

This paper contains 10 sections, 11 equations, 7 figures, 2 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Reactive Pushing System Overview
  3. The Robot Platform
  4. Capacitive Touch Sensor
  5. Reactive Pushing Strategy (RPS)
  6. Experiments and Results
  7. The System Parameters
  8. The Simulation Experiments
  9. The Real-world Experiments
  10. Discussion and Conclusion

Figures (7)

  • Figure 1: Snapshots demonstrating the robot pushing a heavy box (25 kg) toward a target location positioned behind it.
  • Figure 2: Picture and electrical schematic of the capacitive sensing system.
  • Figure 3: High-level scheme of the proposed framework.
  • Figure 4: The set of objects pushed in the simulation experiments: a) a 20 kg box with dimensions of 40 $\times$ 40 $\times$ 80 cm, b) a 5 kg box with dimensions of 45 $\times$ 45 $\times$ 60 cm, and a 10 kg cylinder positioned on top of the rear left corner with a radius of 10 cm and height of 30 cm to simulate asymmetric mass distribution, and c) a 25 kg cylinder with a radius of 25 cm and height of 70 cm.
  • Figure 5: Results of the simulation experiments with a) RPS, b) NPS, and c) APS Pushing_corridor. The plots depict the average minimum distance, denoting the distance between the contact point where the object touches the robot and the goal location, for both sets of friction across every target position and object. The results strongly demonstrate the superior performance of our RPS especially in comparison to NPS and APS.
  • ...and 2 more figures