AgniNav: Configuration-Driven Cross-Embodiment Local Planning for Robot Navigation

2026-06-09

AI Summary

This work proposes a configuration-driven local navigation framework that decouples visual navigation from specific robot embodiments, enabling cross-platform deployment without retraining. By jointly conditioning perception and planning modules on a unified collision envelope parameterized by height, front/rear length, and half-width, the approach achieves shared configurability across diverse morphologies. The perception module employs a height-conditioned monocular-to-pseudo-LiDAR network, while the planner incorporates a dimension-aware local policy based on a four-parameter safety envelope. Evaluated on Turtlebot2, Unitree Go2, and AE K1 robots, the method attains success rates of 39/40, 18/20, and 18/20 respectively, with minimal collisions, and runs in real time at 30 Hz on a Jetson Orin platform.

Abstract

Monocular local navigation is attractive for lightweight robots, but existing vision-based policies often couple perception to a specific body, camera height, and footprint, making transfer from wheeled bases to legged platforms dependent on retraining or active depth hardware. This paper introduces AgniNav, a configuration-driven local navigation framework that standardizes cross-embodiment transfer at the collision-envelope level. Each robot is specified by a measurable four-parameter safety envelope: collision-relevant height, front length, rear length, and half width. The height parameter conditions an image-to-scan network to predict a one-dimensional, collision-relevant pseudo-laserscan from a monocular color image, while the remaining footprint parameters configure a dimension-aware local planner for collision checking. Training uses height-conditioned column-minimum scan labels generated from paired color-depth data, allowing the same image to supervise different safety envelopes without collecting robot-specific data. To the best of our knowledge, AgniNav is the first monocular local-navigation framework that jointly conditions perception and planning on a shared collision-envelope configuration for zero-retraining deployment across wheeled, quadruped, and humanoid platforms. Real-robot experiments on a Turtlebot2, Unitree Go2, and Accelerated Evolution K1 achieve 39/40, 18/20, and 18/20 successes with 0/40, 1/20, and 2/20 collisions, respectively, while running at 30 Hz on Jetson Orin.

Problem

• cross-embodiment
• monocular navigation
• collision envelope
• zero-retraining
• local planning

Innovation

• configuration-driven
• cross-embodiment
• monocular navigation
• collision envelope
• zero-retraining