Published May 7, 2021 | Version v1
Thesis Open

Strategy for an Autonomous Behavior that Guarantees a Qualitative Fine Adjustment at the Target Pose of a Collaborating Mobile Robot

  • 1. Hochschule Luzern

Description

Flexible automation technologies often require an accuracy in the millimeter and milliradian range at different target positions. This thesis introduces a strategy that leverages a pose graph-based localization approach to reduce positioning errors at target poses of collaborating mobile robots. Using the precise localization from the pose graph we propose an odometry-based method to improve the accuracy at target locations, thus avoiding manual teach-in of reference scans for ICP that can degenerate in dynamic environments over time.
As a basis for this strategy, an extensible software architecture for socially acceptable navigation of mobile robots is designed and implemented using ROS components. This architecture proposal includes a DevOps toolchain to automate the process of software testing, building and delivery to different robotic platforms or the cloud. Finally, we evaluate the proposed method in a industrial setting achieving a position error below 25 mm in 92.7% and a rotation error below 1.5° in 93.9% of the tests. Although the advances shown in this thesis couldn’t be statistically tested due to skewed data distributions and some possible outliers, the improvement in both trueness and precision underlines the gains in terms of accuracy positioning.

Files

MT_Kai Waelti.pdf

Files (18.1 MB)

Name Size Download all
md5:79833d5a5db06f80d5debbf19304d528
18.1 MB Preview Download