Low-Cost Air Quality Sensing towards Smart Homes

Round 1

Reviewer 1 Report

The aim of review was well described and organized

In the below sentences,

we found that machine learning-based predictive modelling techniques are superior than~

you need to modify this sentence to include specific superior points. For example, processing time or data accuracy or adaptable to vulnerable conditions or other pros.

Sentences in abstract are little vague to derive and summarize your own findings from the research.

In line 45, spell check

In line 150, 4. sensor technology,

The authors explain well about the sensors` requirements for the smart home applications, but one important function appears to be missing. As you may know, selective sensing is as important as the sensitivity. A highly sensitive sensor can detect ppb level of contaminants, it becomes useless if it can not recognize specific pollutant, especially from mixtures or sometimes identify between humidity and smoke. So, I suggest to include sentences regarding selective sensing for the reliable sensing.

In line 213, very high-selective MOx gas sensors can pick up selected VOCs against a high background of interfering gases

This sentence does not correspond to the metal oxide based LCS and also conflicts with the cross-sensitivity characteristics in gas mixtures of MOx sensor in following sentences. I suggest to remove this part to prevent the confusion to readers.

line 215, it is suggested to write in concise sentences like below or similar simple expression such as poor recovery.

it has a difficulty to fully recover to the initial status under experimental condition changes or exposure to high concentration of target gases

In addition, following comments are suggested to be added in 4.2 chapter. One of drawbacks of MOx gas sensor is heating. Basically, heating is necessary for all MOx gas sensors due to its intrinsic electron conduction sensing principle. So, it limits its applications open to explosive environment and also degrades sensing ability as time goes by.

Space between line 622 and 623

In conclusion part:

Authors insist that controlled laboratory conditions are necessary for the reliable tests for LCSs. However, practical conditions of smart home is not as controlled as those were tested. Therefore, both accurate sensor development and serial data techniques including collection by spatio-mapping, processing and prediction are important to obtain meaningful data, which reflect indoor environment correctly. So, it is suggested to address a guide approach to show the indoor environment as accurate as possible.

It would be better to include the suggestion about future research approach how to fill the gap between sensing technology and smart home. Consequent goal of indoor air sensing is to maintain optimum level of indoor environment by actively operating ventilation system and reducing down the level of indoor pollutants.

Author Response

Reviewer #1

• The aim of review was well described and organized

Response

Thanks for calling our publication “well described and organized”.

• In the below sentences, “we found that machine learning-based predictive modelling techniques are superior than~”, you need to modify this sentence to include specific superior points. For example, processing time or data accuracy or adaptable to vulnerable conditions or other pros.

Response

The sentence is revised as “The review also showed the potential of using machine learning technique for predicting spatio-temporal IAQ in LCS networked-systems.”.

• Sentences in abstract are little vague to derive and summarize your own findings from the research.

Response

The abstract is revised to address the reviewer’ comment.

• In line 45, spell check.

Response

Thanks, corrected.

• In line 150, 4. sensor technology, the authors explain well about the sensors` requirements for the smart home applications, but one important function appears to be missing. As you may know, selective sensing is as important as the sensitivity. A highly sensitive sensor can detect ppb level of contaminants, it becomes useless if it can not recognize specific pollutant, especially from mixtures or sometimes identify between humidity and smoke. So, I suggest to include sentences regarding selective sensing for the reliable sensing.

Response

Thanks for the suggestion, the sentence is revised as “(i) be sensitive and selective to target pollutants for reliable sensing relevant to indoor environments that pose health risks to occupants”.

• In line 213, very high-selective MOx gas sensors can pick up selected VOCs against a high background of interfering gases. This sentence does not correspond to the metal oxide based LCS and also conflicts with the cross-sensitivity characteristics in gas mixtures of MOx sensor in following sentences. I suggest to remove this part to prevent the confusion to readers.

Response

This sentence was removed to prevent the confusion to readers.

• line 215, it is suggested to write in concise sentences like below or similar simple expression such as poor recovery. It has a difficulty to fully recover to the initial status under experimental condition changes or exposure to high concentration of target gases

Response

Thanks for the suggestion, the suggestion is used in Table 2, which reads as “Poor recovery to achieve initial status under a change in experimental condition or exposure to a high concentration of target gases”.

• In addition, following comments are suggested to be added in 4.2 chapter. One of drawbacks of MOx gas sensor is heating. Basically, heating is necessary for all MOx gas sensors due to its intrinsic electron conduction sensing principle. So, it limits its applications open to explosive environment and also degrades sensing ability as time goes by.

Response

Thanks for the suggestion, but it is unrealistic to consider a heated MOx sensor could cause an explosion in smart buildings. For this reason, we exclude this limitation from the study.

Regarding the degradation of sensing stability as time goes by, an entry in Table 2 is revised as “Instability over time”.

• Space between line 622 and 623

Response

Thanks, corrected.

In conclusion part:

• Authors insist that controlled laboratory conditions are necessary for the reliable tests for LCSs. However, practical conditions of smart home is not as controlled as those were tested. Therefore, both accurate sensor development and serial data techniques including collection by spatio-mapping, processing and prediction are important to obtain meaningful data, which reflect indoor environment correctly. So, it is suggested to address a guide approach to show the indoor environment as accurate as possible.

Response

A controlled laboratory condition inside lab creates various environmental conditions and pollution concentrations, representing indoor environments, for thoroughly assessing LCSs before deployment. In-field co-location, spatio-mapping, processing and prediction are the next steps to obtain meaningful data after deployment. This information discussed under this comment is reflected in the revised manuscript to avoid confusing the readers.

• It would be better to include the suggestion about future research approach how to fill the gap between sensing technology and smart home. Consequent goal of indoor air sensing is to maintain optimum level of indoor environment by actively operating ventilation system and reducing down the level of indoor pollutants.

Response

The future section is revised as “The efforts by the above-mentioned projects along with the support of ongoing research activities on transforming homes into smart homes using air quality sensors could result in appreciable health benefits to smart home occupants.”.

Reviewer 2 Report

A brief summary:

The authors provide a review of a range of studies that investigate the implementation of low-cost sensor technology in an indoor environment.  These studies are described and presented in this review in the context of a ‘smart home’ designed to assess and predict occupant exposure to indoor air quality.

Broad comments:

As the article is written now, I think it contains a great deal of useful information across a range of relevant topics.  However, I think the article would benefit from more clarify in the synthesis of these ideas and recommendations.

I think this work would be made more meaningful and would better accomplish the goals identified by the authors through the addition and clear communication of a series of logical frameworks (which could be communicated through flowcharts or diagrams) that are supported by cited studies, relating user goals and applications to appropriate planning and design toward implementation of a successful ‘smart’ indoor sensor network in terms of

• sensor selection
• sensor calibration
• deployment
• data QA/QC
• efficacy of an indoor smart sensors system to predict occupant exposure

The application area of LCS networks in this article is indoor environments, but outdoor air quality standards, measurements, and sensor networks are also discussed.   Near the beginning of the introduction, the authors need to explain connections between indoor and outdoor air quality, and how these are relevant in this article. 

The authors point out a gap in the literature, indicating that they: ‘comprehensively assessed the pivotal aspect of sensor selection and performance assessment of LCSs before/during use’. (lines 87/88)  It makes sense to deploy sensors for specific pollutants of concern in different locations in a home, as the authors describe, but what about issues around sensor sensitivity to humidity, temperature, and confounding gas species?  How should these details inform sensor selection?  I don’t think this article effectively addresses or fills this gap.

 The authors describe how their work can help fill a gap in connection with the need of  ‘high-resolution spatio-temporal coverage for long-term deployment.’ (Line 89).  Did the authors collect and/or analyze data toward this end? I didn’t see this being specifically addressed in the article.

The authors highlight the importance of data reliability and QA/QC, but I don’t think the literature review includes sufficient coverage of this topic. For example, which sensor calibration techniques are most effective and why?  What about ‘field normalization’ also called ‘field calibration’?

Specific comments:

Line 45:  summarizes common indoor air pollutants?

Line 49:  PM should be singular

Line 50:  counted as ‘an’ air pollutant

Line 54:  before introducing Table 1, I think it would be helpful to discuss the relevance of outdoor air pollution in the context of indoor air quality.

Line 54:  instead of ‘reference values’ do the authors mean unhealthy exposure thresholds?

Line 54:  Table 1:  units need to be clearly labeled/specified for all numbers presented in this table.

Line 58:  ‘engaging in people’s everyday lives’?  needs clarification

Line 60:  high spatial resolution and real-time air pollution data?

Line 61:  what is meant by ‘large networks’ ?

Line 64:  ‘USD<2500’ needs clarification….  I think the authors mean ‘less than $2500 (USD)’?

Line 65:  ‘around few USD’ this phrase and the rest of the sentence need clarification. 

Lines 69/70:  ‘environment-friendly environment’ meaning is unclear….

Line 70:  the outcomes of what?

Line 80: ‘particulate’??  wrong word?

Line 93:  is the goal to make a home smart or to characterize air quality in a home?  Isn’t air quality characterization just one aspect of what a smart home could be?

Line 301:  ‘least’ attention?  Not sure what is meant here.

Line 567:  ‘’subsequent’?  maybe instead:  ‘integrated’

Line 577:  ‘IAQ standards’?  the authors highlight just before, that ‘IAQ remains mostly unregulated’. These statements seem to negate each other.  Is the problem lack of regulation?  Is the problem occupant exposure?  We could argue both….

Line 578:  what is meant by ‘different disciplines’?  needs clarification

Line 579:  which ‘components’?  what type of diversity? 

Line 580:  ‘IAQ demands?’  do the authors mean ‘measurement challenges’?

Author Response

Reviewer #2

A brief summary:

The authors provide a review of a range of studies that investigate the implementation of low-cost sensor technology in an indoor environment. These studies are described and presented in this review in the context of a ‘smart home’ designed to assess and predict occupant exposure to indoor air quality. Broad comments:

• As the article is written now, I think it contains a great deal of useful information across a range of relevant topics. However, I think the article would benefit from more clarify in the synthesis of these ideas and recommendations.

Response

Thanks, we have carefully revisited the manuscript and incorporated comments/suggestions from reviewers.

• I think this work would be made more meaningful and would better accomplish the goals identified by the authors through the addition and clear communication of a series of logical frameworks (which could be communicated through flowcharts or diagrams) that are supported by cited studies, relating user goals and applications to appropriate planning and design toward implementation of a successful ‘smart’ indoor sensor network in terms of
• sensor selection
• sensor calibration
• deployment
• data QA/QC
• efficacy of an indoor smart sensors system to predict occupant exposure

Response

A summary of the main steps towards this goal is added into a process diagram as shown in Figure 1.

• The application area of LCS networks in this article is indoor environments, but outdoor air quality standards, measurements, and sensor networks are also discussed. Near the beginning of the introduction, the authors need to explain connections between indoor and outdoor air quality, and how these are relevant in this article.

Response

A sentence “Additionally, IAQ could be affected by local outdoor air pollutants, which can ingress into indoor environments (see Table 1).” is added to show the connection between outdoor and indoor air quality. The authors are in agreement with the reviewer regarding the importance of ambient air quality and considering outdoor/indoor air pollutants as highlighted in Section 3.

• The authors point out a gap in the literature, indicating that they: ‘comprehensively assessed the pivotal aspect of sensor selection and performance assessment of LCSs before/during use’. (lines 87/88) It makes sense to deploy sensors for specific pollutants of concern in different locations in a home, as the authors describe, but what about issues around sensor sensitivity to humidity, temperature, and confounding gas species? How should these details inform sensor selection? I don’t think this article effectively addresses or fills this gap.

Response

Thanks for highlining this point, the sentence is removed to avoid confusing the readers. In addition, the issues related to cross-sensitivities are reviewed in Section 4.

• The authors describe how their work can help fill a gap in connection with the need of ‘high-resolution spatio-temporal coverage for long-term deployment.’ (Line 89). Did the authors collect and/or analyze data toward this end? I didn’t see this being specifically addressed in the article.

Response

Thanks for pointing out of this, the sentence is revised as “(ii) review and summarise the optimal deployment strategies of LCSs within a domestic context”.

• The authors highlight the importance of data reliability and QA/QC, but I don’t think the literature review includes sufficient coverage of this topic. For example, which sensor calibration techniques are most effective and why? What about ‘field normalization’ also called ‘field calibration’?

Response

The field normalisation (or field calibration) is a necessary part of QA/QC. This study highlights the use of affordable tools to calibrate and evaluate the performance of LCSs before and after deployment. They could improve the readings due to testing sensors under conditions representing the deployed environments. The authors further clarified their view in the revised manuscript.

Specific comments:

• Line 45: summarizes common indoor air pollutants?

Response

It is corrected as suggested.

• Line 49: PM should be singular.

Response

Corrected as suggested.

• Line 50: counted as ‘an’ air pollutant.

Response

Corrected as suggested.

• Line 54: before introducing Table 1, I think it would be helpful to discuss the relevance of outdoor air pollution in the context of indoor air quality.

Response

A sentence “Additionally, IAQ could be affected by local outdoor air pollutants, which can ingress into indoor environments (see Table 1).” is added to bring the connection between these two.

• Line 54: instead of ‘reference values’ do the authors mean unhealthy exposure thresholds?

Response

It is changed to “The unhealthy exposure thresholds defined for the common indoor and outdoor air pollutants”.

• Line 54: Table 1: units need to be clearly labeled/specified for all numbers presented in this table.

Response

The units are assigned to each pollutant as commented.

• Line 58: ‘engaging in people’s everyday lives’? needs clarification

Response

It is replaced by “in everyday life”.

• Line 60: high spatial resolution and real-time air pollution data?

Response

It is revised as “…with the prospect of delivering real-time air pollution data through sensor networks”.

• Line 61: what is meant by ‘large networks’?

Response

It is revised as sensor networks.

• Line 64: ‘USD<2500’ needs clarification…. I think the authors mean ‘less than $2500 (USD)’?

Response

Yes, it is revised as “less than $2500 (USD)”.

• Line 65: ‘around few USD’ this phrase and the rest of the sentence need clarification. Response

This phrase is removed in the revision.

• Lines 69/70: ‘environment-friendly environment’ meaning is unclear….

Response

The sentence is revised as “Smart homes can enable an adaptable living environment, e.g. …”

• Line 70: the outcomes of what?

Response

The revision reads as “The outcomes of these projects could contribute in …”.

• Line 80: ‘particulate’?? wrong word?

Response

It is deleted to bring clarity to the sentence.

• Line 93: is the goal to make a home smart or to characterize air quality in a home? Isn’t air quality characterization just one aspect of what a smart home could be?

Response

We agree about the comment, thereby, the sentence is revised as “…for making homes smart with respect to air quality”.

• Line 301: ‘least’ attention? Not sure what is meant here.

Response

“less attention” is the right term in here as corrected.

• Line 567: ‘’subsequent’? maybe instead: ‘integrated’

Response

It is replaced by “integrated” as suggested.

• Line 577: ‘IAQ standards’? the authors highlight just before, that ‘IAQ remains mostly unregulated’. These statements seem to negate each other. Is the problem lack of regulation? Is the problem occupant exposure? We could argue both….

Response

It reads better in the revised form “Smart homes equipped with air quality LCSs and integrated processing/predicting tools can offer a healthy environment to occupants”.

• Line 578: what is meant by ‘different disciplines’? needs clarification,

Response

This is revised along with other comments as you can see in the next comment.

• Line 579: which ‘components’? what type of diversity?

Response

The sentence is revised as “Although technologies in this field are continuously ongoing, emerging knowledge among the researchers in different fields is sparse, and smart home components are considered separately due to diversity in the research field.”.

• Line 580: ‘IAQ demands?’ do the authors mean ‘measurement challenges’?

Response

It is replaced with “indoor measurement challenges” as suggest

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Authors have revised correctly as reviewers suggested and commented