Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

ClutterNav: Gradient-Guided Search for Efficient 3D Clutter Removal with Learned Costmaps

Navin Sriram Ravie, Keerthi Vasan M, Bijo Sebastian

TL;DR

ClutterNav is presented, a novel decision-making framework that can identify the next best object to be removed so as to access a target object in a given clutter, while minimising stack disturbances.

Abstract

Dense clutter removal for target object retrieval presents a challenging problem, especially when targets are embedded deep within densely-packed configurations. It requires foresight to minimize overall changes to the clutter configuration while accessing target objects, avoiding stack destabilization and reducing the number of object removals required. Rule-based planners when applied to this problem, rely on rigid heuristics, leading to high computational overhead. End-to-end reinforcement learning approaches struggle with interpretability and generalizability over different conditions. To address these issues, we present ClutterNav, a novel decision-making framework that can identify the next best object to be removed so as to access a target object in a given clutter, while minimising stack disturbances. ClutterNav formulates the problem as a continuous reinforcement learning task, where each object removal dynamically updates the understanding of the scene. A removability critic, trained from demonstrations, estimates the cost of removing any given object based on geometric and spatial features. This learned cost is complemented by integrated gradients that assess how the presence or removal of surrounding objects influences the accessibility of the target. By dynamically prioritizing actions that balance immediate removability against long-term target exposure, ClutterNav achieves near human-like strategic sequencing, without predefined heuristics. The proposed approach is validated extensively in simulation and over real-world experiments. The results demonstrate real-time, occlusion-aware decision-making in partially observable environments.

ClutterNav: Gradient-Guided Search for Efficient 3D Clutter Removal with Learned Costmaps

TL;DR

ClutterNav is presented, a novel decision-making framework that can identify the next best object to be removed so as to access a target object in a given clutter, while minimising stack disturbances.

Abstract

Dense clutter removal for target object retrieval presents a challenging problem, especially when targets are embedded deep within densely-packed configurations. It requires foresight to minimize overall changes to the clutter configuration while accessing target objects, avoiding stack destabilization and reducing the number of object removals required. Rule-based planners when applied to this problem, rely on rigid heuristics, leading to high computational overhead. End-to-end reinforcement learning approaches struggle with interpretability and generalizability over different conditions. To address these issues, we present ClutterNav, a novel decision-making framework that can identify the next best object to be removed so as to access a target object in a given clutter, while minimising stack disturbances. ClutterNav formulates the problem as a continuous reinforcement learning task, where each object removal dynamically updates the understanding of the scene. A removability critic, trained from demonstrations, estimates the cost of removing any given object based on geometric and spatial features. This learned cost is complemented by integrated gradients that assess how the presence or removal of surrounding objects influences the accessibility of the target. By dynamically prioritizing actions that balance immediate removability against long-term target exposure, ClutterNav achieves near human-like strategic sequencing, without predefined heuristics. The proposed approach is validated extensively in simulation and over real-world experiments. The results demonstrate real-time, occlusion-aware decision-making in partially observable environments.

Paper Structure

This paper contains 24 sections, 7 equations, 11 figures, 5 tables.

Table of Contents

  1. Introduction
  2. Methodology
  3. Preliminaries
  4. Simulation Environment and Clutter Generation
  5. Instance Segmentation on Point Clouds
  6. Reinforcement Learning for Cost Estimation
  7. Feature Extraction
  8. Why Handcrafting?
  9. Removability Critic
  10. Removability Scores
  11. Reward Function Design
  12. Gradient Guided Object Importance Estimation
  13. Exact Influence Gradient Formulation
  14. Integrated Gradients
  15. Complete Importance Score
  16. ...and 9 more sections

Figures (11)

  • Figure 1: Hardware setup used for experimental validation.
  • Figure 2: Overview of the proposed ClutterNav framework. (Top) Real-world setup showing a UR5e arm with a 3F140 gripper executing clutter removal to retrieve a target object. (Bottom) System pipeline combining offline removability cost learning from simulation and online gradient-guided planning for efficient and safe target retrieval.
  • Figure 3: Sim Setup, (a) Clutter on the table (b) Raw point cloud (c) Observable Point Cloud (d) Cuboid fitting on point cloud
  • Figure 4: Lightweight SAC Architecture displaying the observable object features (green), passed through the Actor after padding (grey), outputting the action (orange) concatenated with the padded feature vector and passed through the critic, giving the removability scores for the observable objects
  • Figure 5: Cost Landscape for our experimental arrangements - top to bottom - Random Clutter, Stack, Wall and Pyramid
  • ...and 6 more figures