cs.RONov 2, 2026

h-NESDF: An Infinitely Differentiable Signed Distance Between Two Sets with Robotic Applications

Authors: Vinicius Mariano Gonçalves, Pedro Antonio de Souza Campos Rodrigues

Organizations: Department of Electrical Engineering, Federal University of Minas Gerais (UFMG), Belo Horizonte, Brazil · Control and Automation Engineering Undergraduate Program at Federal University of Minas Gerais (UFMG), Belo Horizonte, Brazil

Abstract

Computing distances between sets is important in many robotic applications, such as motion planning and control. In many applications, derivatives of the distance are needed as well. It is well known that the Euclidean distance is not always differentiable, which create issues to algorithms that need a derivative. In this paper, we introduce NESDF, the Normal Euclidean Signed Distance Function between two general (with mild regularity assumptions but possibly non-convex) sets. This metric is still not differentiable, but it allows us to introduce a generalization, hh-NESDF, which has a smoothing parameter hh such that the metric is infinitely differentiable for all h>0h>0 and that reduces to the NESDF when hh goes to 00. The metric is amenable to GPU parallelization, and thus can be computed efficiently. We show extensive simulation results to evaluate its viability in motion planning.

Figures & tables

Explore similar work

CardsList
  1. OREN: Octree Residual Network for Real-Time Euclidean Signed Distance Mapping

    Oct 21, 2025Zhirui Dai, Qihao Qian, Tianxing Fan +13D MappingImplicit Neural Representations

  2. CSSDF-Net: Safe Motion Planning Based on Neural Implicit Representations of Configuration Space Distance Field

    Mar 19, 2026Haohua Chen, Yixuan Zhou, Yifan Zhou +1Safe Motion PlanningRobot Collision Avoidance