A Novel 3D Complex Welding Seam Tracking Method in Symmetrical Robotic MAG Welding Process Using a Laser Vision Sensing

Round 1

Reviewer 1 Report

The requirements during the welding processes that led to this research are not explained.

The author should explane the reasons leading to the need for seam tracking technique to guide the robot during arc welding.

The authors should explain the innovative results that were added since publication:

A Novel Seam Tracking Technique with a Four-Step Method and Experimental Investigation of Robotic Welding Oriented to Complex Welding Seam

Gong Zhang 1,2 , Yuhang Zhang 1,3, Shuaihua Tuo 1,3, Zhicheng Hou 1,*, Wenlin Yang 1 , Zheng Xu 1 , Yueyu Wu 1 , Hai Yuan 1 and Kyoosik Shin

Journal Sensors

The contribution of the actirvity would be more innovative if the authors could introduce some of the results mentioned in the conclusions:

“The possible directions of future work comprise the following ideas: In order to 361 achieve better tracking performance for 3D complex welding seams, multi-sensor infor- 362 mation fusion technology, reinforcement learning and deep neural network may be ap- 363 plied to the image processing flow of welding seams, and the quality inspection of com- 364 plex weldments. For instance, X-ray image and convolutional neural network (CNN) 365 could be used for the detection and recognition of the weld seam defects; Convolution 366 filter and deep reinforcement learning (RL) could be combined to localize the weld fea- 367 ture point in each welding image; Visual tracking and object detection based on a deep 368 learning (DL) framework could be proposed to address the problem of low welding pre- 369 cision caused by possible external environmental disturbances.”

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

In my opinion, the article under review describes the solution to an urgent problem facing the specialists of the welding industry at the present time, namely the implementation of curved welds. The developed technology can be used in additive technologies for arc growth of metal curvilinear products. The article has a finished look and contains all the necessary structural elements. The support of five grants proves the importance of this research work.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

In this article the authors address the problem of three-dimensional tracing of complex weld seams. In particular, a four-step weld bead tracking system is proposed: segmented scanning, combined filtering, characteristic point extraction and weld path planning. In order to improve joint tracking accuracy, a combined filtering technique is used to correct the data to reduce the effects of flash, data distortion and noise on the weld surface. Finally, a spatial weld path is achieved by weld path planning. The experimental aspect is addressed with experimental investigations on two-dimensional (2D) and three-dimensional (3D) curved weld seams. The results obtained demonstrate that the proposed method can form a complete weld path.The authors show a good tracking method with satisfactory experimental results. It certainly can provide, as the authors themselves claim, a reference for research into high-precision seam tracking and automatic welding.

The authors show a good tracking method with satisfactory experimental results. It certainly can provide, as the authors themselves claim, a reference for research into high-precision seam tracking and automatic welding.

The work is well presented, both in the description, and how it is presented. The formulas are well laid out even if they have not been checked by me. The figures should be improved a little in clarity, especially no. 8 which is quite confusing. Furthermore, I believe that the subject is interesting but does not present a particular attraction in the reference scientific literature. Ultimately, I believe that the work can be published as presented, with the suggestion to integrate the references with the following papers:

1) Cammarata, A., Sinatra, R., Maddio, P.D.

A two-step algorithm for the dynamic reduction of flexible mechanisms (2019). Mechanisms and Machine Science, 66, pp. 25-32. DOI: 10.1007/978-3-030-00365-4_4

 

2) Cammarata, A., Sinatra, R., Maddìo, P.D.

Static condensation method for the reduced dynamic modeling of mechanisms and structures. (2019) Archive of Applied Mechanics, 89 (10), pp. 2033-2051. DOI: 10.1007/s00419-019-01560-x

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Dear authors, thank you for the revisions. My suggestion is to add more sensors in order to have data to feed a machine learning algorithm that can contribute to the development of a physical-virtual welding process.

 

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf