Occluding the Solution Space: Planner-Agnostic Adversarial Attacks on Tolerance-Aware Manipulation

2026-07-04

AI Summary

Existing adversarial attacks on motion planning rely on precise pose targets and require planner-in-the-loop queries, limiting their generalizability. This work proposes a planner-agnostic adversarial attack framework that shifts the attack objective from exact pose accuracy to task-level feasibility, evaluating system robustness by masking feasible solution spaces. The key innovation lies in introducing a kinematic occupancy heatmap to characterize workspace capability, combined with offline modeling and online constrained maximum coverage optimization to efficiently deploy adversarial obstacles under geometric constraints. Experiments demonstrate that the proposed method significantly outperforms baseline approaches in both simulation and real-world scenarios, reliably inducing planning failures with higher computational efficiency.

Abstract

Adversarial attacks on motion planning are crucial for evaluating and quantifying the intrinsic robustness of robotic manipulation. However, existing approaches are typically limited by restrictive exact-pose objectives and their reliance on planner-in-the-loop queries. To address these limitations, we propose a planner-agnostic attack framework for tolerance-aware manipulation. Our approach shifts the evaluation paradigm to task-level feasibility over goal regions, efficiently inserting adversarial obstacles without requiring oracle access to the victim system. Offline, we characterize the robot's intrinsic workspace capabilities via a kinematic occupancy heatmap, which encodes the density of feasible trajectories and robustness priors without invoking a specific planner. Online, we formulate the attack as a budgeted maximum-coverage optimization, strategically deploying obstacles subject to explicit geometric constraints to occlude the solution space. Extensive experiments across simulation and real-world scenarios demonstrate that our method reliably induces planning failures, significantly outperforming planner-in-the-loop baselines in both computational efficiency and attack efficacy.

Problem

• adversarial attacks
• motion planning
• planner-agnostic
• tolerance-aware manipulation
• solution space occlusion

Innovation

• planner-agnostic
• tolerance-aware manipulation
• kinematic occupancy heatmap
• adversarial obstacle insertion
• maximum-coverage optimization