Adapting Learning by Demonstration for Robot Based Part Feeding Applications

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

Abstract

Feeding parts in automated manufacturing processes lack the adaptation flexibility to rapid production changes, which are typical for small to medium enterprises. In this paper we present a novel application approach for feeding parts, which can be incorporated in small and medium production facilities. The approach relies on a scoop like tool that is attached to a collaborative robot. With the adoption of learning by demonstration methods, part feeding skills can be set-up fast and easy directly on the factory floor by a non-expert user. The approach exploits dynamic movement primitives as an underlying method for representation of trajectories. In the execution phase, we use admittance control with switching force control parameters, defined during the demonstration, for reliable execution of the part feeding skill in contact with the environment. The proposed approach is tested on on several part feeding cases with a Universal Robot UR10e.

Original languageEnglish
Title of host publicationProceedings of the 2020 IEEE/SICE International Symposium on System Integration, SII
PublisherIEEE
Publication date2020
Pages954-959
ISBN (Print)978-1-7281-6668-1
ISBN (Electronic)978-1-7281-6667-4
DOIs
Publication statusPublished - 2020
Event2020 IEEE/SICE International Symposium on System Integration (SII) - Honolulu, United States
Duration: 12. Jan 202015. Jan 2020

Conference

Conference2020 IEEE/SICE International Symposium on System Integration (SII)
Country/TerritoryUnited States
CityHonolulu
Period12/01/202015/01/2020

Fingerprint

Dive into the research topics of 'Adapting Learning by Demonstration for Robot Based Part Feeding Applications'. Together they form a unique fingerprint.

Cite this