Basic Study on Rapid Estimation of Machining Time Based on AI with Two-Dimensional Data (Trihedral Figures)

Abstract—Machining time estimation is necessary for production scheduling, and it is important to answer the delivery date and price instantly when receiving an order from a customer. Currently, machining time is estimated by computer-aided manufacturing (CAM) system, but machining time estimation takes time for a numerical control (NC) program creation and machining simulation. Therefore, it is necessary to instantaneously estimate machining time. So, in this paper, we develop a system to estimate machining time instantaneously using artificial intelligence (AI), the input to the AI system was a trihedral figure of the shape to be removed, and the output was machining time (in intervals of 15 minutes). In this paper, we used convolutional neural network (CNN) which is a kind of AI and effective for image recognition, and estimated the machining time. Then, we created the shape to be removed by creating the required shape (machine parts) and the material shape (rectangular prisms) arbitrarily as the machining data, and estimated the machining time from the removal volume, and constructed the data set. An evaluation experiment was performed to allow AI to train 1082 images of the trihedral figure of the shape to be removed and confirm the estimation accuracy of the machining time. As a result of conducting evaluation experiments, it was possible to obtain a machining time estimation result within 15 minutes of prediction error in all 70evaluation data. In this paper, the outline of the proposed method, the method of creating the machining data of self-made, and the method of constructing the optimal CNN are described in order, and finally, the results of the evaluation experiment are summarized.

Index Terms—Artificial intelligence, image processing, machining time estimation, trihedral figure.

Hiroki Takizawa is with the School of Integrated Design Engineering, Keio University, Yokohama, Japan (e-mail: takizawa@ddm.sd.keio.ac.jp).
Hideki Aoyama is with Department of System Design Engineering, Keio University, Yokohama, Japan (e-mail: haoyama@sd.keio.ac.jp).
Song Cheol Won is with the UEL Corporation, Japan (e-mail: cheolwon.song@excel.co.jp).

[PDF]

Cite: Hiroki Takizawa, Hideki Aoyama, and Song Cheol Won, "Basic Study on Rapid Estimation of Machining Time Based on AI with Two-Dimensional Data (Trihedral Figures)," International Journal of Materials, Mechanics and Manufacturing vol. 7, no. 6, pp. 235-239, 2019.

Copyright © 2019 by the authors. This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited (CC BY 4.0).