Model Free Calibration of Wheeled Robots Using Gaussian Process

TL;DR

This paper introduces a Gaussian Process-based non-parametric method for calibrating wheeled robots using onboard sensor data, which is flexible, does not require external ground truth, and can handle unknown drive configurations.

Contribution

It presents a novel, general calibration approach that learns the entire sensor/robot motion model from odometry data without external ground truth, including a linear approximation for efficiency.

Findings

Accurately calibrates robots with unmodeled effects.

Does not require external motion capture systems.

Demonstrates flexibility across different robot configurations.

Abstract

Robotic calibration allows for the fusion of data from multiple sensors such as odometers, cameras, etc., by providing appropriate relationships between the corresponding reference frames. For wheeled robots equipped with camera/lidar along with wheel encoders, calibration entails learning the motion model of the sensor or the robot in terms of the data from the encoders and generally carried out before performing tasks such as simultaneous localization and mapping (SLAM). This work puts forward a novel Gaussian Process-based non-parametric approach for calibrating wheeled robots with arbitrary or unknown drive configurations. The procedure is more general as it learns the entire sensor/robot motion model in terms of odometry measurements. Different from existing non-parametric approaches, our method relies on measurements from the onboard sensors and hence does not require the ground truth information from external motion capture systems. Alternatively, we propose a computationally efficient approach that relies on the linear approximation of the sensor motion model. Finally, we perform experiments to calibrate robots with un-modelled effects to demonstrate the accuracy, usefulness, and flexibility of the proposed approach.