Dear Nguyen Hong Son,T ran Trung Hieu

Reference is made to your article:

SELECTION OF WELDING ROBOT BY MULTI-CRITERIA DECISION MAKING METHOD

I have read your article and my comments are:

1. In page 2 you say “Therefore, in order to have a solid basis for asserting that a certain option is the best, it is necessary to rank those options with multiple sets of weights determined by different methods”

This is inexact. You don’t need for ranking options multiple sets of weights by different methods. Many people run different methods on the same problem, normally, using the same set of weights, unless they are performing simulations. My question is: What for? What is the purpose of that exercise? Nothing is gained.

2- Page 2 “The use of two MCDM methods with different characteristics (MARCOS and PSI) along with four different weighting methods was expected to yield a conclusion with the highest reliability”

This is not so. You can use any number of methods, with the same or different set of weights, normally yielding different rankings. But even if rankings have high correlation, it does not assure high reliability of results.

3. Page 7 “That allows us to say that determining the best alternative does not de[1]pend on the MCDM method and the weighting method used (at least with the methods used in this study)”

I am afraid that I disagree. Not all methods are good for any kind of problem, and not all methods are capable of mathematical modelling certain characteristics of a scenario.

4. In page 6 “At that time, the decision-making of multiple criteria to choose the best option must consider these criteria further. This is a job to do in the future in choosing robots”

Not really. All characteristics of a problem whatever they are, must be considered at the same time and from the very beginning. Normally, there are not several stages, unless the results in cases where the first feed a second.

5. Page 6 “The data normalization and weighting methods used in this study can only be applied when the criteria are in quantitative format. These methods cannot be applied when the criteria are in qualitative form such as ease of use, warranty coverage, customer preferences, etc. This is the limitation of this study”

In this case it appears that your example is incomplete, because normally there is a mix of quantitative and qualitative criteria, and all of them important.

Hope this helps

Nolberto Munier