Localizing semi-static objects in AMR applications: A comparison of sensors and algorithms

Other Titles
Lokalisierung von Semi-statischen Objekten in AMR Applikationen: Ein Vergleich von Sensoren und Algorithmen
Publikationstyp
Journal Article
Date Issued
2024-10-30
Sprache
English
Author(s)
Renz, Matthias Johannes 
Technische Logistik W-6  
Rose, Hendrik Wilhelm 
Technische Logistik W-6  
Ziegenbein, Justin  orcid-logo
Braun, Philipp Maximilian  orcid-logo
Technische Logistik W-6  
TORE-URI
https://tore.tuhh.de/handle/11420/53488
Journal
Logistics journal / Proceedings  
Volume
2024
Issue
20
Citation
Logistics Journal 2024 (20): (2024-10-30)
Publisher DOI
10.2195/lj_proc_renz_en_202410_01
Scopus ID
2-s2.0-85214536709
Publisher
Wissenschaftliche Gesellschaft fur Technische Logistik
Autonomous mobile robots (AMR) have substantial impact on the automation of logistics processes like last mile delivery. In order to securely enter or interact with objects, accurate positioning of the object in the robot’s maps is required. If a large object is semi-static, occupies a large part of the surrounding and is previously known, different localization approaches can be used for positioning the object relative to the AMR. This contribution compares approaches for the position of semi-static objects in an AMR’s map such as AMCL, ICP and AprilTag detection using the robot’s LiDAR and cam-eras. It also develops an evaluation scheme to rate the approaches qualitatively and quantitatively to choose the most appropriate solution for the use case in hand. Based on the rating scheme, AprilTag localization proved to be the best performer for a last mile delivery robot entering a carrier vehicle.
Subjects
Autonomous Mobile Robots (AMR) | Localization | real-world research | Semistatic Obstacles
DDC Class
388: Transportation
620: Engineering