Leimbach, Lars (2024) Acquisition and Analysis of Surface Processing Demonstrations for Robotic Execution. Masterarbeit, Technical University of Munich.
PDF
- Nur DLR-intern zugänglich
146MB |
Kurzfassung
In the realm of industrial robotics, the integration of human expertise into robotic systems represents a significant frontier. This thesis presents a comprehensive study analyzing expert demonstration techniques for surface polishing and their application in robotic operations. The primary aim is to enhance the precision and efficiency of robots in performing tasks traditionally executed by skilled humans, specifically focusing on sanding and polishing processes in manufacturing. The research encapsulates the development of a sophisticated data acquisition system that records detailed trajectories of expert movements, including position, orientation, and forces. These trajectories are analyzed to extract nuanced patterns and strategies, which are then translated into robotic instructions. A novel framework is proposed that leverages this analysis to enable robots to replicate human-like efficiency in surface processing tasks. Experimental results are shown on the robot. The study includes 21 participants and it also explores the human efficiency regarding covering an object. The most frequent and efficient pattern was the snake pattern. The findings underscore the potential of integrating human dexterity into robotic systems, paving the way for broader applications in automated manufacturing and beyond. This work contributes to the academic field by providing a detailed data acquisition and analysis approach. Additionally, it offers practical insights for future work.
elib-URL des Eintrags: | https://elib.dlr.de/206110/ | ||||||||
---|---|---|---|---|---|---|---|---|---|
Dokumentart: | Hochschulschrift (Masterarbeit) | ||||||||
Titel: | Acquisition and Analysis of Surface Processing Demonstrations for Robotic Execution | ||||||||
Autoren: |
| ||||||||
Datum: | 16 Juni 2024 | ||||||||
Erschienen in: | Acquisition and Analysis of Surface Processing Demonstrations for Robotic Execution | ||||||||
Open Access: | Nein | ||||||||
Seitenanzahl: | 73 | ||||||||
Status: | veröffentlicht | ||||||||
Stichwörter: | robotic surface processing, polishing, grinding, sanding, smoothing, Learning from Demonstration, Programming by Demonstration, imitation learning, robot skill, parameter extraction | ||||||||
Institution: | Technical University of Munich | ||||||||
Abteilung: | Department of Informatics | ||||||||
HGF - Forschungsbereich: | Luftfahrt, Raumfahrt und Verkehr | ||||||||
HGF - Programm: | Raumfahrt | ||||||||
HGF - Programmthema: | Robotik | ||||||||
DLR - Schwerpunkt: | Raumfahrt | ||||||||
DLR - Forschungsgebiet: | R RO - Robotik | ||||||||
DLR - Teilgebiet (Projekt, Vorhaben): | R - Synergieprojekt Factory of the Future Extended | ||||||||
Standort: | Oberpfaffenhofen | ||||||||
Institute & Einrichtungen: | Institut für Robotik und Mechatronik (ab 2013) | ||||||||
Hinterlegt von: | Eiband, Thomas | ||||||||
Hinterlegt am: | 02 Sep 2024 12:41 | ||||||||
Letzte Änderung: | 02 Sep 2024 12:41 |
Nur für Mitarbeiter des Archivs: Kontrollseite des Eintrags