ConFusion: Sensor Fusion for Complex Robotic Systems using Nonlinear Optimization

TL;DR

ConFusion is an open-source, modular sensor fusion framework for robotics that uses nonlinear optimization within a moving horizon estimator, offering greater flexibility and scalability than traditional filtering methods.

Contribution

It introduces a flexible, scalable sensor fusion framework using nonlinear optimization, outperforming Kalman filter-based methods in robotic applications.

Findings

ConFusion outperforms iterated extended Kalman filter in visual-inertial tracking.

It demonstrates effective whole-body sensor fusion on a mobile manipulator.

The framework is open-source and adaptable to various robotic sensor configurations.

Abstract

We present ConFusion, an open-source package for online sensor fusion for robotic applications. ConFusion is a modular framework for fusing measurements from many heterogeneous sensors within a moving horizon estimator. ConFusion offers greater flexibility in sensor fusion problem design than filtering-based systems and the ability to scale the online estimate quality with the available computing power. We demonstrate its performance in comparison to an iterated extended Kalman filter in visual-inertial tracking, and show its versatility through whole-body sensor fusion on a mobile manipulator.