Internet of Things (IoT) Based Indoor Air Quality Sensing and Predictive Analytic—A COVID-19 Perspective
, Syed Mohammad Hassan Zaidi 1, Muhammad Zeeshan Shakir 2,*, Uferah Shafi 1, Muhammad Moeez Malik 1, Ayesha Haque 3, Sadaf Mumtaz 3 and Syed Ali Raza Zaidi 4Round 1
Reviewer 1 Report
Dear authors,
I believe that the topic of your research is very actual, of high relevance in the current pandemic times, and in some extent, of interest to the audience of the journal. My suggestion is to motivate and elaborate more on the theoretical arguments of your research.
Some minor English revisions are needed, due to some typos (for example, "Predicitve" in figure 2).
Author Response
Please see the attachment
Author Response File: 
Author Response.pdf
Reviewer 2 Report
- Suggest addressing the scope of the manuscript to focus on why and how to improve the air quality monitor approach for indoor air quality, and highlight the method contributions.
- One important factor for IoT devices is the power management, what's the power consumption of the air quality device, and what's the battery life time?
- The classification and prediction parts are using existing machine learning models, suggest the author to justify the novelties of the classification and predictive approaches.
- Figure 3, suggest showing the hardware architecture instead of PCB, PCB design can be an appendix. Architecture is more informative and showing the full design of the hardware.
- Can you add content about measure calibration for each type of the sensor? And how is the ground truth obtained, and what's the comparison between the proposed hardware's measurements and the ground truth?
Author Response
Please see the attachment
Author Response File: Author Response.pdf
Reviewer 3 Report
The paper presents an experiment where data about air quality dimensions were collected with several sensors, and that data was use to generate some prediction methods to be able to predict future air quality. Also a web environment and an app were developed to support the system.
The topic is very interesting and up-to-date.
Following my comments and suggestions:
1 - Abbreviations should be explained before its use, for instance HVAC and AC in the introduction;
2 - References are up-to-date, nevertheless, there are too many references that are weblinks. In my opinion, some scientific works from journals and conferences in the area should be included
3 - (line 127) I do prefer "research" instead of "investigation"
4 - Novelties and contributions should be clearly stated in the introduction
5 - Structure for the rest of the article should be presented at the end of the introduction
6 - In section 3.1, why do you select those pollutants and not others?
7 - (line 276) DO you mean figures 2-3?
8 - Figure 5 comes before figure 4. Also figure 11 comes before figure 10. Please correct.
9 - (line 285) Do you mean Figures 4, 5, and 6?
10 - (line 289) Which lab are you referring to? Why was it chosen for the experiment?
11 - From figure 16, a question arises: where is the target variable, i.e., the Class? In addition, what are you really predicting? Or, is this a time series prediction? These aspects needs more explanations from the authors
12 - In the graphics from section 5, I suggest that the authors use different types of lines to facilitate those who print the paper in black and white
13 - Limitations of this work should be presented in the introduction
14 - Last, but not the least: would anything be different in the system without COVID-19? Don't you think that this is just a motivation to develop your work? Isn't it a little far-fetched to consider COVID-19 in the title?
Author Response
Please see the attachment
Author Response File: Author Response.pdf
Reviewer 4 Report
Review of Electronics 1008591 Internet of things based IAQ
Abstract
Line 12. IAQ can affect more than COVID and those with pulmonary disorders. For example, PM can affect heart disease and neurological outcomes. The authors might just add: “and other health problems” to this sentence.
Line 17. The term “NN” is not defined.
The abstract does not state how the data were obtained and whether the sensors were precise and accurate. It also does not state which contaminants were measured and whether performance varied for different types of contaminants. IAQ is not a single variable.
Line 34. It seems odd not to mention vaccines at all here.
Line 46. It should be Harvard University, not the University of Harvard
Line 57. This sentence seems to imply that relative humidity causes COPD. Sentence should be reworded.
Line 104. It is not clear why the authors chose these contaminants, but not others, such as carbon monoxide, VOCs, formaldehyde and others.
Line 123. I am not aware that electron mobility spectrometers and X-Ray fluorescence Spectrometers have been used to measure IAQ.
Line 130. The term “hob” is not defined
Line 219. Why were these contaminants selected instead of others?
Line 294. The method of validation is not adequately described
Figure 6. It is not clear why certain levels have been defined as being “safe” and the figure does not explain what the numbers on it mean.
Figures 7-10. It is not clear why these figures are needed. It is well known that contaminants vary in concentration over time.
Line 362. This is the first mention of missing data. Why were they missing and how many missing data points are there?
Line 544. The conclusion is simply a restatement of points already made.
The World Health Organization Indoor Air Quality Guidelines are not referenced.
This article requires extensive improvements in the English and is far too repetitious and too long.
Author Response
Please see the attachment
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
- Figure 7 - Figure 14, what's the sensor sampling rate? And can you make the monitoring time longer to show one day's cycle?
- And can you improve the visual presentation quality of the Figure 7 - Figure 14?
- Figure 17, Figure 19, the predicted value shows a phase shift? Can you add explanations of the shift, and is the shift corrected in the accuracy comparison?
- Figure 17, Figure 19, can you put timestamp on x-axis instead if T1,...T49?
- For prediction, how accurate for future 1 hour, 24 hours prediction?
Author Response
Please see the attachment
Author Response File: Author Response.pdf
Reviewer 4 Report
No comments on this revision, except that the English is still deficient in the revision.
Author Response
Please see the attachment
Author Response File: Author Response.pdf
Round 3
Reviewer 2 Report
Thank you for addressing all the questions!
