Adaptive acquisition planning for visual inspection in remanufacturing using reinforcement learning

Publikation: Beiträge in ZeitschriftenZeitschriftenaufsätzeForschungbegutachtet

Standard

Adaptive acquisition planning for visual inspection in remanufacturing using reinforcement learning. / Kaiser, Jan Philipp; Gäbele, Jonas; Koch, Dominik et al.
in: Journal of Intelligent Manufacturing, 2024.

Publikation: Beiträge in ZeitschriftenZeitschriftenaufsätzeForschungbegutachtet

Harvard

APA

Vancouver

Kaiser JP, Gäbele J, Koch D, Schmid J, Stamer F, Lanza G. Adaptive acquisition planning for visual inspection in remanufacturing using reinforcement learning. Journal of Intelligent Manufacturing. 2024. doi: 10.1007/s10845-024-02478-0

Bibtex

@article{d58a74016da6416bba723cb71f349ec8,
title = "Adaptive acquisition planning for visual inspection in remanufacturing using reinforcement learning",
abstract = "In remanufacturing, humans perform visual inspection tasks manually. In doing so, human inspectors implicitly solve variants of visual acquisition planning problems. Nowadays, solutions to these problems are computed based on the object geometry of the object to be inspected. In remanufacturing, however, there are often many product variants, and the existence of geometric object models cannot be assumed. This makes it difficult to plan and solve visual acquisition planning problems for the automated execution of visual inspection tasks. Reinforcement learning offers the possibility of learning and reproducing human inspection behavior and solving the visual inspection problem, even for problems in which no object geometry is available. To investigate reinforcement learning as a solution, a simple simulation environment is developed, allowing the execution of reproducible and controllable experiments. Different reinforcement learning agent modeling alternatives are developed and compared for solving the derived visual planning problems. The results of this work show that reinforcement learning agents can solve the derived visual planning problems in use cases without available object geometry by using domain-specific prior knowledge. Our proposed framework is available open source under the following link: https://github.com/Jarrypho/View-Planning-Simulation.",
keywords = "Acquisition planning, Inspection, Reinforcement learning, Remanufacturing, View planning, Engineering",
author = "Kaiser, {Jan Philipp} and Jonas G{\"a}bele and Dominik Koch and Jonas Schmid and Florian Stamer and Gisela Lanza",
note = "Publisher Copyright: {\textcopyright} The Author(s) 2024.",
year = "2024",
doi = "10.1007/s10845-024-02478-0",
language = "English",
journal = "Journal of Intelligent Manufacturing",
issn = "0956-5515",
publisher = "Springer Netherlands",

}

RIS

TY - JOUR

T1 - Adaptive acquisition planning for visual inspection in remanufacturing using reinforcement learning

AU - Kaiser, Jan Philipp

AU - Gäbele, Jonas

AU - Koch, Dominik

AU - Schmid, Jonas

AU - Stamer, Florian

AU - Lanza, Gisela

N1 - Publisher Copyright: © The Author(s) 2024.

PY - 2024

Y1 - 2024

N2 - In remanufacturing, humans perform visual inspection tasks manually. In doing so, human inspectors implicitly solve variants of visual acquisition planning problems. Nowadays, solutions to these problems are computed based on the object geometry of the object to be inspected. In remanufacturing, however, there are often many product variants, and the existence of geometric object models cannot be assumed. This makes it difficult to plan and solve visual acquisition planning problems for the automated execution of visual inspection tasks. Reinforcement learning offers the possibility of learning and reproducing human inspection behavior and solving the visual inspection problem, even for problems in which no object geometry is available. To investigate reinforcement learning as a solution, a simple simulation environment is developed, allowing the execution of reproducible and controllable experiments. Different reinforcement learning agent modeling alternatives are developed and compared for solving the derived visual planning problems. The results of this work show that reinforcement learning agents can solve the derived visual planning problems in use cases without available object geometry by using domain-specific prior knowledge. Our proposed framework is available open source under the following link: https://github.com/Jarrypho/View-Planning-Simulation.

AB - In remanufacturing, humans perform visual inspection tasks manually. In doing so, human inspectors implicitly solve variants of visual acquisition planning problems. Nowadays, solutions to these problems are computed based on the object geometry of the object to be inspected. In remanufacturing, however, there are often many product variants, and the existence of geometric object models cannot be assumed. This makes it difficult to plan and solve visual acquisition planning problems for the automated execution of visual inspection tasks. Reinforcement learning offers the possibility of learning and reproducing human inspection behavior and solving the visual inspection problem, even for problems in which no object geometry is available. To investigate reinforcement learning as a solution, a simple simulation environment is developed, allowing the execution of reproducible and controllable experiments. Different reinforcement learning agent modeling alternatives are developed and compared for solving the derived visual planning problems. The results of this work show that reinforcement learning agents can solve the derived visual planning problems in use cases without available object geometry by using domain-specific prior knowledge. Our proposed framework is available open source under the following link: https://github.com/Jarrypho/View-Planning-Simulation.

KW - Acquisition planning

KW - Inspection

KW - Reinforcement learning

KW - Remanufacturing

KW - View planning

KW - Engineering

UR - http://www.scopus.com/inward/record.url?scp=85202164754&partnerID=8YFLogxK

U2 - 10.1007/s10845-024-02478-0

DO - 10.1007/s10845-024-02478-0

M3 - Journal articles

AN - SCOPUS:85202164754

JO - Journal of Intelligent Manufacturing

JF - Journal of Intelligent Manufacturing

SN - 0956-5515

ER -

DOI