Tracking Snake-like Robots in the Wild Using Only a Single Camera

TL;DR

This paper presents a novel method for tracking snake-like robots in complex environments using only a single camera, combining differentiable rendering with Kalman filtering for accurate pose and joint estimation.

Contribution

It introduces a marker-less, single-camera tracking approach that estimates robot pose and joint angles simultaneously, even in unstructured and moving camera scenarios.

Findings

Achieved 0.05 m accuracy in base position localization

Estimated joint angles within 6 degrees of error

Validated in both stationary and moving camera setups

Abstract

Robot navigation within complex environments requires precise state estimation and localization to ensure robust and safe operations. For ambulating mobile robots like robot snakes, traditional methods for sensing require multiple embedded sensors or markers, leading to increased complexity, cost, and increased points of failure. Alternatively, deploying an external camera in the environment is very easy to do, and marker-less state estimation of the robot from this camera's images is an ideal solution: both simple and cost-effective. However, the challenge in this process is in tracking the robot under larger environments where the cameras may be moved around without extrinsic calibration, or maybe when in motion (e.g., a drone following the robot). The scenario itself presents a complex challenge: single-image reconstruction of robot poses under noisy observations. In this paper, we address the problem of tracking ambulatory mobile robots from a single camera. The method combines differentiable rendering with the Kalman filter. This synergy allows for simultaneous estimation of the robot's joint angle and pose while also providing state uncertainty which could be used later on for robust control. We demonstrate the efficacy of our approach on a snake-like robot in both stationary and non-stationary (moving) cameras, validating its performance in both structured and unstructured scenarios. The results achieved show an average error of 0.05 m in localizing the robot's base position and 6 degrees in joint state estimation. We believe this novel technique opens up possibilities for enhanced robot mobility and navigation in future exploratory and search-and-rescue missions.