Reinforcement Learning-based Curriculum Learning Design and Experiment of Grid Sortation System in Smart Factory

Journal of Advanced Technology Research, Vol. 4, No. 2, pp. 6-15, Dec. 2019
10.11111/JATR.2020.4.2.006, Full Text:
Keywords: Reinforcement Learning, Curriculum Learning, Smart Factory, Grid Sortation System
Abstract

Artificial intelligence(AI) has recently changed the paradigm in various industries and societies, but state-of-the-art AI has not shown immediate results in manufacturing. In other words, practical methodologies that differentiate from conventional approaches are needed at Industry 4.0. Manufacturing works with a number of physical combinations of hardware. Thus, in the real world, it is impossible to repeat tens of millions of times without the intervention of engineers because of physical wear, failure, etc. In addition, manufacturing is very difficult to collect and label vast amounts of data for learning. One way to overcome these two limitations is to reproduce a very similar environment to reality in a simulation and then use reinforcement learning. After developing a grid classification system and designing reinforcement learning, we apply curriculum learning and show that efficient behavior control is possible.


Statistics
Show / Hide Statistics

Statistics (Cumulative Counts from September 1st, 2017)
Multiple requests among the same browser session are counted as one view.
If you mouse over a chart, the values of data points will be shown.


Cite this article
[IEEE Style]
H. Choi, G. Hwang, H. Lim, J. Kim and Y. Han, "Reinforcement Learning-based Curriculum Learning Design and Experiment of Grid Sortation System in Smart Factory," Journal of Advanced Technology Research, vol. 4, no. 2, pp. 6-15, 2019. DOI: 10.11111/JATR.2020.4.2.006.

[ACM Style]
Ho-Bin Choi, Gyu-Young Hwang, Hyun-Kyo Lim, Ju-Bong Kim, and Youn-Hee Han. 2019. Reinforcement Learning-based Curriculum Learning Design and Experiment of Grid Sortation System in Smart Factory. Journal of Advanced Technology Research, 4, 2, (2019), 6-15. DOI: 10.11111/JATR.2020.4.2.006.