Simpler learning of robotic manipulation of clothing by utilizing DIY smart textile technology
- Author
- Andreas Verleysen (UGent) , Thomas Holvoet, Remko Proesmans (UGent) , Cedric Den Haese and Francis wyffels (UGent)
- Organization
- Abstract
- Deformable objects such as ropes, wires, and clothing are omnipresent in society and industry but are little researched in robotics research. This is due to the infinite amount of possible state configurations caused by the deformations of the deformable object. Engineered approaches try to cope with this by implementing highly complex operations in order to estimate the state of the deformable object. This complexity can be circumvented by utilizing learning-based approaches, such as reinforcement learning, which can deal with the intrinsic high-dimensional state space of deformable objects. However, the reward function in reinforcement learning needs to measure the state configuration of the highly deformable object. Vision-based reward functions are difficult to implement, given the high dimensionality of the state and complex dynamic behavior. In this work, we propose the consideration of concepts beyond vision and incorporate other modalities which can be extracted from deformable objects. By integrating tactile sensor cells into a textile piece, proprioceptive capabilities are gained that are valuable as they provide a reward function to a reinforcement learning agent. We demonstrate on a low-cost dual robotic arm setup that a physical agent can learn on a single CPU core to fold a rectangular patch of textile in the real world based on a learned reward function from tactile information.
- Keywords
- deformable object manipulation, smart textile, low-cost, reinforcement learning
Downloads
-
applsci-10-04088.pdf
- full text (Published version)
- |
- open access
- |
- |
- 7.96 MB
Citation
Please use this url to cite or link to this publication: http://hdl.handle.net/1854/LU-8664963
- MLA
- Verleysen, Andreas, et al. “Simpler Learning of Robotic Manipulation of Clothing by Utilizing DIY Smart Textile Technology.” APPLIED SCIENCES-BASEL, vol. 10, no. 12, 2020, doi:10.3390/app10124088.
- APA
- Verleysen, A., Holvoet, T., Proesmans, R., Den Haese, C., & wyffels, F. (2020). Simpler learning of robotic manipulation of clothing by utilizing DIY smart textile technology. APPLIED SCIENCES-BASEL, 10(12). https://doi.org/10.3390/app10124088
- Chicago author-date
- Verleysen, Andreas, Thomas Holvoet, Remko Proesmans, Cedric Den Haese, and Francis wyffels. 2020. “Simpler Learning of Robotic Manipulation of Clothing by Utilizing DIY Smart Textile Technology.” APPLIED SCIENCES-BASEL 10 (12). https://doi.org/10.3390/app10124088.
- Chicago author-date (all authors)
- Verleysen, Andreas, Thomas Holvoet, Remko Proesmans, Cedric Den Haese, and Francis wyffels. 2020. “Simpler Learning of Robotic Manipulation of Clothing by Utilizing DIY Smart Textile Technology.” APPLIED SCIENCES-BASEL 10 (12). doi:10.3390/app10124088.
- Vancouver
- 1.Verleysen A, Holvoet T, Proesmans R, Den Haese C, wyffels F. Simpler learning of robotic manipulation of clothing by utilizing DIY smart textile technology. APPLIED SCIENCES-BASEL. 2020;10(12).
- IEEE
- [1]A. Verleysen, T. Holvoet, R. Proesmans, C. Den Haese, and F. wyffels, “Simpler learning of robotic manipulation of clothing by utilizing DIY smart textile technology,” APPLIED SCIENCES-BASEL, vol. 10, no. 12, 2020.
@article{8664963,
abstract = {{Deformable objects such as ropes, wires, and clothing are omnipresent in society and industry but are little researched in robotics research. This is due to the infinite amount of possible state configurations caused by the deformations of the deformable object. Engineered approaches try to cope with this by implementing highly complex operations in order to estimate the state of the deformable object. This complexity can be circumvented by utilizing learning-based approaches, such as reinforcement learning, which can deal with the intrinsic high-dimensional state space of deformable objects. However, the reward function in reinforcement learning needs to measure the state configuration of the highly deformable object. Vision-based reward functions are difficult to implement, given the high dimensionality of the state and complex dynamic behavior. In this work, we propose the consideration of concepts beyond vision and incorporate other modalities which can be extracted from deformable objects. By integrating tactile sensor cells into a textile piece, proprioceptive capabilities are gained that are valuable as they provide a reward function to a reinforcement learning agent. We demonstrate on a low-cost dual robotic arm setup that a physical agent can learn on a single CPU core to fold a rectangular patch of textile in the real world based on a learned reward function from tactile information.}},
articleno = {{4088}},
author = {{Verleysen, Andreas and Holvoet, Thomas and Proesmans, Remko and Den Haese, Cedric and wyffels, Francis}},
issn = {{2076-3417}},
journal = {{APPLIED SCIENCES-BASEL}},
keywords = {{deformable object manipulation,smart textile,low-cost,reinforcement learning}},
language = {{eng}},
number = {{12}},
pages = {{10}},
title = {{Simpler learning of robotic manipulation of clothing by utilizing DIY smart textile technology}},
url = {{http://doi.org/10.3390/app10124088}},
volume = {{10}},
year = {{2020}},
}
- Altmetric
- View in Altmetric
- Web of Science
- Times cited: