RAM: Reachability Across Morphologies

2026-06-08

AI Summary

Existing methods for modeling robotic reachable workspaces often suffer from slow computation, low accuracy, or limited applicability to specific robot configurations. This work proposes a morphology-conditioned implicit neural representation that serves as a fast, differentiable surrogate model for pose reachability, enabling the first universal, efficient, and self-collision-aware reachability modeling across diverse robot morphologies. By integrating implicit neural networks, large-scale forward kinematics sampling, morphological parameterization, and differentiable rendering, the method achieves an F1 score of 86%—a 14% improvement over baseline approaches—while operating at nanosecond-scale inference times, representing a three-order-of-magnitude speedup. Furthermore, it accelerates configuration and trajectory optimization tasks by factors of 10× and 100×, respectively.

Abstract

Many stages of the robotic lifecycle, from morphology synthesis to operation, rely fundamentally on the reachable workspace. However, current methods for approximating workspaces are slow, imprecise, or tied to a single morphology. We introduce Reachability Across Morphologies (RAM): a morphology-conditioned, implicit neural representation that acts as a fast, differentiable surrogate for pose reachability, generalising to unseen morphologies while inherently accounting for self-collisions. To train RAM, we publish a large-scale dataset of $3\cdot10^{10}$ samples generated solely from forward kinematics. Experiments show that our model achieves an $ F_1$-score of $86\%$ at nanosecond inference, outperforming the baseline by $14\%$ while reducing inference time by three orders of magnitude. We further demonstrate speed-ups of one and two orders of magnitude for gradient-based morphology and trajectory optimisation, respectively. Website: https://timwalter.github.io/ram.

Problem

• reachable workspace
• morphology generalization
• robot kinematics
• self-collision
• pose reachability

Innovation

• implicit neural representation
• morphology-conditioned reachability
• self-collision awareness
• differentiable surrogate
• large-scale forward kinematics dataset