A Tool Condition Monitoring System Based on Low-Cost Sensors and an IoT Platform for Rapid Deployment

Round 1

Reviewer 1 Report

verse 370: symbol of steel used in research UNE F114 should be additionally given in ISO standard

verse 375: is 250 m/mina – should be: 250 m/min

An erroneous indication by the tool condition monitoring system for a new tool may relate to the running-in stage. During this time, the cutting forces increase and thus the current drawn by the motor and the level of sound generated by the cutting process. An experienced operator will ignore a system error. The implementation of a system for automatic tool change in cnc machines may pose a threat.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

After reviewing the paper, here are my comments:

1- There are several similarity indices to be utilized, what is the main reason to select equation (1) as the similarity index in the presented article?

2- What is the threshold value for the similarity index to make a decision about if the tested equipment is worn or new?

3- Please enrich the conclusion and support it with data.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 3 Report

The authors have presented a nice idea for TCM. The paper is well-written and organized. A comprehensive literature review is presented. However, there are some ambiguities regarding when the tool is considered a worn tool and how this effect was incorporated into the current approach. The following are the comments to improve the manuscript.

Major comments

1.      One of the major shortcomings in the presented work is that the authors have not defined what is meant by a worn tool. Which criterion is used to define a worn tool? 0.3 mm flank wear or other standard was considered to determine a worn tool. Figure 7 shows that the authors have used rake face wear (crater wear) as the reference for tool wear. But this must be clarified in the manuscript. Also, if rake face wear is used as a reference, then it also needs to be elaborated why the author did not use an easy-to-measure and more widely used flank face wear criterion.

2.      The worn tool's sound signal and the current signal must be presented.

3.      The reference/calibration signal in Figure 10 is acquired when a new tool is used. However, it is well known that tool wear is a progressive phenomenon. As the tool wear increases, the spindle load (force or current) and the machining sound also change. How was this effect considered in the current TCM approach where the reference single is only from a new tool? This means that the current TCM model will declare the signal from only a 0.25 mm worn tool as rejected or WORN tool, while in common practice, the flank wear of up to 0.3 mm is allowed, and then the tool is declared as a worn tool.

4.      It would be more helpful if the authors could provide time on the abscissa for Figure 8 and especially for figure 9 instead of samples.

5.      Line 438 and 439, 10 samples are too small to decide on the classification accuracy of the system.

6.      The conclusions must be rewritten; at the moment, the conclusion section is very general and looks like an abstract.

Minor comments

7.      Line 348, "Secondly, a set of experiments is conducted to create the data…" How many experiments were conducted? How many experiments are the minimum requirement? Did these experiments continue until the tool was completely worn out? Please elaborate on these points in section 2.5.

8.      Line 358, "The result of the model in terms of accuracy and other metrics is 358 sent back to the app for operator's validation." It would be better to highlight what is meant by model accuracy and what are the other metrics. Also, it is not clear how the operator validates the data.

9.      Lines 377 to 382 are methodology, not results. Better to relocate these to the methodology sections.

10.  The authors have not provided any conclusive remarks in section 3.1 regarding line 387, "In order to prove that the current and sound sensor can provide enough information for TCM purposes". It needs to be elaborated on why the authors think that current and sound sensors are enough for TCM.

11.  Some typos need to be corrected, such as in line 23 "few hours is required", line 272 "…with no complex…", line 375 "m/mina", line 472 "CNC controlled", 

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Round 2

Reviewer 3 Report

The authors have satisfactorily revised the manuscript. I recommend accepting the manuscript.