Adaptation of a wheel loader automatic bucket filling neural network using reinforcement learningShow others and affiliations
2020 (English)In: 2020 International Joint Conference on Neural Networks (IJCNN), IEEE, 2020, article id 20563Conference paper, Published paper (Refereed)
Abstract [en]
Bucket-filling is a repetitive task in earth-moving operations with wheel-loaders, which needs to be automated to enable efficient remote control and autonomous operation. Ideally, an automated bucket-filling solution should work for different machine-pile environments, with a minimum of manual retraining. It has been shown that for a given machine-pile environment, a time-delay neural network can efficiently fill the bucket after imitation-based learning from 100 examples by one expert operator. Can such a bucket-filling network be automatically adapted to different machine-pile environments without further imitation learning by optimization of a utility or reward function? This paper investigates the use of a deterministic actor-critic reinforcement learning algorithm for automatic adaptation of a neural network in a new pile environment. The algorithm is used to automatically adapt a bucket-filling network for medium coarse gravel to a cobble-gravel pile environment. The experiments presented are performed with a Volvo L180H wheel-loader in a real-world setting. We found that the bucket-weights in the novel pile environment can improve by five to ten percent within one hour of reinforcement learning with less than 40 bucket-filling trials. This result was obtained after investigating two different reward functions motivated by domain knowledge.
Place, publisher, year, edition, pages
IEEE, 2020. article id 20563
Series
International Joint Conference on Neural Networks (IJCNN), ISSN 2161-4393, E-ISSN 2161-4407
Keywords [en]
Adaptation models, Neural networks, Learning (artificial intelligence), Predictive models, Task analysis, Trajectory, Wheels, reinforcement learning, imitation learning, bucket filling, wheel loader, automation, construction
National Category
Other Electrical Engineering, Electronic Engineering, Information Engineering Control Engineering
Research subject
Cyber-Physical Systems; Control Engineering
Identifiers
URN: urn:nbn:se:ltu:diva-81003DOI: 10.1109/IJCNN48605.2020.9206849ISI: 000626021402015Scopus ID: 2-s2.0-85093819698OAI: oai:DiVA.org:ltu-81003DiVA, id: diva2:1472526
Conference
2020 International Joint Conference on Neural Networks (IJCNN), 19-24 July, 2020, Glasgow, United Kingdom
Funder
Vinnova, 2017-01958The Kempe Foundations, SMK-1429
Note
ISBN för värdpublikation: 978-1-7281-6926-2, 978-1-7281-6927-9
2020-10-012020-10-012023-09-04Bibliographically approved