Sustainable Scheduling of an Automatic Pallet Changer System by Multi-Objective Evolutionary Algorithm with First Piece Inspection
Round 1
Reviewer 1 Report
A hard three-phase multiobjective model is presented, which is important in practice.
A metahuristic approach is proposed to solve the problem, and the efficiency is shown by tests.
I have significant problem with the definition of the model, which should be definitely improved.
I list some detailed remarks.
1. Page 3, Problem Description, line 4. Here ther are two tasks, task O_i and task O_j. The latter is denoted by j, please fix it.
2. Page 4, last paragraph, regarding "According to the standard expression (α |β| γ) of the scheduling problem proposed by Pinedo et al. in [21]...". This was proposed much earlier, by
R.L. Graham, E.L. Lawler, J.K. Lenstra and A.H.G. Rinnooy Kan, Optimization and approximation in deterministic sequencing and scheduling: a survey, Annals of Operations Research 5 (1979), 287-326.
3. Regareding the model, (2)-(12). Here first the input must be given. What is the number of tasks, what are the parameters. What is the number of pallets? What is hte meaning of the notations? For example, what is c_j,k in (7)? What is mu_j in (9)? Why fa_j is chosen so? What is the connection between c_j and c_j,k? How Equations 10 ensures that two operations are not overlapping if both of them are assigned on the same pallet? (A space is missing just after thissentence.) Note that this equation must hold for any j,k pair, but there is not any k in this equation! The explanations for (10) and (11) are the same. Why? What is P_k,g in (12)?
Note also that the makespan minimization is usually denoted as C_max, like on page 12 (and not as MS).
4. Figure 6 is nice! I like also Fig 8.
5. Please give vertical lines into Table 3 between the data for the different methods.
6. I found the Notation just before the References. It should be put before the mathematical model. I propose also to write simply "task" instead of "batch task", as any task is a batch task here, so there is no need to write batch.
Author Response
Please refer to the attached file.
Author Response File: 
Author Response.docx
Reviewer 2 Report
This paper proposes a multi-objective evolutionary algorithm considering pass rate of the first piece inspection in APC task scheduling. I think it is a valuable paper because it utilizes a real-world application and suggests an optimization algorithm from a new point of view.
However, following issues need to be revised and supplemented. (a file is attached)
1) From the perspective of sustainability, the contribution of this paper is unclear.
2) There are many missing notations. Also, visibility is low because the notation is at the end of the paper. There are a lot of typos in the text, so overall correction is required.
3) An overall review of the mathematical model presented in p5 is required. The distinction between task and operation is not clear, and there is a lack of explanation. It is also unclear whether the problem under consideration is to schedule, i) tasks in a single operation, ii) operations, iii) tasks in multiple operations. If it is the case iii), you should separate the subscripts or describe this situation clearly.
4) What does the integer in the chromosome's gene mean? In order to consider RCFPI, the distinction between Pallet A and Pallet B is essential. What is the process?
5) In Figure 1, it may be helpful to show the connection with Next two tasks further. In addition, is it necessary to change pallet during processing of the tasks in the Normal phase?
6) The SDP of Eq (1) plays an important role in the algorithm. However, since SDP occurs only in the first piece inspection phase, I cannot be sure that the probability calculation of Eq (1) is correct. A more complex process would be required to calculate the exact probability.
7) If the scheduling procedure proceeds as shown in Figure 1, idling times are always small when the operation with higher pass rate in the two pallets is performed first. So, if this is the case, does the complex evolutionary algorithm required to solve this scheduling problem? In fact, as shown in the Gantt charts in Figure 11, mixing higher pass rate operation into a combination is critical to improving performance.
7) It is not clear why the simplified APC evaluation model on page 8 is needed. Even with the non-simplified model, the computation time seems to be short enough.
8) The proposed evolutionary algorithm and performance metrics are typical. A clear explanation of the contribution of this paper is required.
9) The boxplots shown in Figure 7 clearly present that the proposed APCEA performs better than other algorithms. However, the trends in Figure 8 show that there is not much difference in performance between algorithms at the termination. It would be worthwhile to show how the absolute magnitude differences of HVR, IGD, and SP are valuable in a real manufacturing environment.
Comments for author File: 
Comments.pdf
Author Response
Please refer to the attached file.
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
In this revised version I see that my previous requests are applied.
I see only some very minor issues to change, I list them below.
Page 4, just after Figure 1. Here we find "fig 1" in the text. Please write with capital letter as Fig.
Page 5, paragraph "The objective of the machining", after "Graham et al.[25]," please put a space.
Also, page 6, in the text with red. After any "Equation(?)" there must be a space.
The much data in the table on pages 17 and 18 are too dense. Not too much can be seen well. I still think that some kind of reorganization of the data would be helpful.
Fig 8 is nice, but please help the reader to tell that in case (a) the higher is the better, and in all other cases the lower is the better.
Author Response
Please refer to the attached file.
Author Response File: Author Response.docx
Reviewer 2 Report
I think it is possible to publish this paper because the authors have solved all the suggestions I proposed.
Author Response
Dear Reviewer:
Thank you for your comments concerning our manuscript entitled “Sustainable Scheduling of an Automatic Pallet Changer System by Multi-Objective Evolutionary Algorithm with First Piece Inspection”.
We would like to again express our appreciations to the reviewer for the careful, conscientious and beneficial comments.
Thank you very much for your work concerning our paper.
Best Regards
Yours sincerely
Qingmiao Liao, Jianjun Yang and Yong Zhou
2019-3-7
