TY - GEN
T1 - Toward Fully Automated Metal Recycling using Computer Vision and Non-Prehensile Manipulation
AU - Han, Shuai D.
AU - Huang, Baichuan
AU - Ding, Sijie
AU - Song, Changkyu
AU - Feng, Si Wei
AU - Xu, Ming
AU - Lin, Hao
AU - Zou, Qingze
AU - Boularias, Abdeslam
AU - Yu, Jingjin
N1 - Funding Information:
1 S. D. Han, B. Huang, C. Song, S. W. Feng, A. Boularias, and J. Yu are with the Department of Computer Science, Rutgers University, Piscataway, NJ, USA. Emails: shuai.han@rutgers.edu, {baichuan.huang, changkyu.song, siwei.feng, abdeslam.boularias, jingjin.yu}@cs.rutgers.edu. 2 S. Ding is with Brown University, Providence, RI, USA. Email: sijie ding@brown.edu. 3 M. Xu is with GEM Co. Ltd., Tianjin, China, Email: xuming@gem.com.cn. 4 H. Lin and Q. Zou are with the School of Engineering, Rutgers University, Piscataway, NJ, USA. Emails: {hlin, qzzou}@soe.rutgers.edu. This work is supported in part by a research contract from HSG (Wuhan) Internet Co..
Publisher Copyright:
© 2021 IEEE.
PY - 2021/8/23
Y1 - 2021/8/23
N2 - Due to inherent irregularities in recyclable materials, sorting valuable metals (e.g., aluminum and copper) via mechanical means is a difficult task resisting full automation. A particularly hard challenge in the domain is the separation of scrap metal pieces with physically attached impurities, which is further complicated by variations in different batches of recyclable materials. In this work, leveraging the latest development in machine learning and robot learning, we develop an image-based sorting system for tackling this challenging task. In addition to delivering a highly accurate deep learning model for reliably distinguishing pure scrap pieces from pieces containing impurities with over 95% precision/recall, we further automate the process of sample preparation, data acquisition/labeling/analysis, and machine learning model training.
AB - Due to inherent irregularities in recyclable materials, sorting valuable metals (e.g., aluminum and copper) via mechanical means is a difficult task resisting full automation. A particularly hard challenge in the domain is the separation of scrap metal pieces with physically attached impurities, which is further complicated by variations in different batches of recyclable materials. In this work, leveraging the latest development in machine learning and robot learning, we develop an image-based sorting system for tackling this challenging task. In addition to delivering a highly accurate deep learning model for reliably distinguishing pure scrap pieces from pieces containing impurities with over 95% precision/recall, we further automate the process of sample preparation, data acquisition/labeling/analysis, and machine learning model training.
UR - http://www.scopus.com/inward/record.url?scp=85117059121&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85117059121&partnerID=8YFLogxK
U2 - 10.1109/CASE49439.2021.9551431
DO - 10.1109/CASE49439.2021.9551431
M3 - Conference contribution
AN - SCOPUS:85117059121
T3 - IEEE International Conference on Automation Science and Engineering
SP - 891
EP - 898
BT - 2021 IEEE 17th International Conference on Automation Science and Engineering, CASE 2021
PB - IEEE Computer Society
T2 - 17th IEEE International Conference on Automation Science and Engineering, CASE 2021
Y2 - 23 August 2021 through 27 August 2021
ER -