Verification of Cyberphysical Systems

Round 1

Reviewer 1 Report

The article is well written and well structured. The topic is attractive and presented in an interesting way, using sometimes colloquial language but therefore it is easy follow.

The cyber-physical systems are currently applied in many domains of our life. Hence, it is important to ensure their high quality and correctness of functionality. The authors propose an approach for formal verification of such system with usage of Lingua Franca (language for programming) and Rebeca (language for model checking).

There are some ways in which the paper could be improved:

1. The authors could provide a little bit more information about the model checking itself and its advantages. This is a common used formal verification technique, but not all readers may be familiar with it.
2. The subchapters should be considered. When there appears subchapter 2.1 – the reader expects then subchapter 2.2, but is does not follow. The same happens for the next main sections (section three & five).
3. Figure 4a should be redrawn so that different transitions have different style. In the current form they differ only in color, and when printing the article in a grey-scale the reader is enforced to look into an electronic version to interpret the transitions correctly. Consider using either colors with various brightness or lines with different style (e.g. thick and thin).
4. The authors write, that the limited number of experiments have been conducted. My question is – is the number of conducted experiments sufficient to draw such conclusions as stated in the article? Do the experiments concern similar size of models?
5. Then performing formal verification with the model checking technique one can be faced with the state explosion problem. Although the authors state that they have to be careful regarding the state space explosion, they do not provide any details how they are solving (or avoiding) the problem.
6. The article is based on a previous conference paper, where some parts look quite similar. Could the authors provide the information what are the detailed differences? There are only two sentences in the article regarding this issue.
7. Simple grammatical errors, e.g. missing a closing bracket in line 509 on page 17.

Author Response

Please see the attached file for our response to the reviewers.

Author Response File: Author Response.pdf

Reviewer 2 Report

I would suggest two topics for further discussion:

1. State space and possible state explosion. The considered problem seems to be simple, and the reachability graph shown in Fig. 4 has 10 nodes (for other considered variants of the model it will be bigger). So, is it possible to evaluate the complexity of the models (in terms of e.g. number of signals and/or parameters related to the logical time intervals) for which the presented approach still can be applied?
2. In the considered case "there is no need to introduce an extra actor to model the environment". That's because the case is simple, but when there are many signals from the enviromnent and many of their combinations are impossible because of the way of functioning of the physical part of the system, introducing of such model seems to be necessary (or at least reasonable). How the approach would look like in such case?

Author Response

Please see the attached file for our response to the reviewers.

Author Response File: Author Response.pdf

Reviewer 3 Report

The article proposes a new formal verification model for cyber-physical systems (CPS) that focuses on the cyber part based on a reliable model of the interface to the physical part.

This method is based on the assumption that the alignment of different timelines during system execution is the responsibility of the underlying platforms. To what extent is this assumption valid?

The integration of the physical and cybernetic systems is a very important question to prove in this type of system. You have to test the entire chain. However in the article due to the assumption made this part is obvious.

This means that the article does not make a novel contribution, it simply focuses on a type of CPS verification test.

It seems to me that the idea of putting existing tools into practice to solve a problem for which it seems they were designed, may not be very helpful. However, the implementation, although it seems simple, is acceptable.

The state of the art must improve, it has very old references from the years 1977, 1990 and 1996. Justify its inclusion or replace them with more modern ones.

The experiment is well explained and illustrates its applicability, although as I say it seems simple and could incorporate more assumptions to verify. You should also better highlight the results. Maybe there should be a section summarizing results.

I think a section on related works is missing and compare with other techniques to see how this proposal differs.

Why did authors choose this particular journal?

What is the novelty of the article?

For all this, I recommends the paper to be accepted with major revision to reinforce the experiment, the state of the art regarding related works and then clarify results, detected bugs and how they have been corrected.

Author Response

Please see the attached file for our response to the reviewers.

Author Response File: Author Response.pdf

Round 2

Reviewer 3 Report

Accept in present form