Indoor Occupancy Sensing via Networked Nodes (2012–2022): A Review
Round 1
Reviewer 1 Report
This article reviews the development of occupancy detection systems for indoor spaces, with a focus on using networked sensor nodes for improved accuracy and coverage. The review covers studies from 2012 to 2022 that use Passive Infrared (PIR) based sensors and compares methods based on Application Desired Parameters (ADPs) such as accuracy and feasible detection area. The article introduces a new metric called "sensor node deployment density" to measure the strength of network-level data filtering and fusion algorithms, which are crucial for robust occupancy estimation. The article provides a standardized insight into occupancy detection pipelines, including data collection and connectivity strategies, sensor models, data fusion, machine learning algorithms, and occupancy estimation algorithms. Finally, the article identifies gaps in past research and recommends the suitability of reviewed methods for different application areas such as Health & Safety or Occupant Comfort.
Overall, this paper needs many improvements on different aspects:
1) The motivation of proposing this review should be well presented in the abstract and introduction.
2) The main contributions of this paper is not clear. There are other occupancy detection review that have published recently. The authors should compare the contributions with the other, such as: Review on occupancy detection and prediction in building simulation; A comprehensive review of approaches to building occupancy detection; Deep and transfer learning for building occupancy detection: A review and comparative analysis; Occupancy detection systems for indoor environments: A survey of approaches and methods; Occupancy detection in non-residential buildings–A survey and novel privacy preserved occupancy monitoring solution; Occupancy detection and localization strategies for demand modulated appliance control in Internet of Things enabled home energy management system
3) To do that please insert a new section called comparison with existing review and then compare the different contriubutions in the text and by adding a table that summarizes the main contributions of each review.
3) The abbreviation list should be inserted at the beginning of the article.
5) Many recent occuapncy detection articles are missing, such as: OccupancySense: Context-based indoor occupancy detection & prediction using CatBoost model; Unobtrusive occupancy and vital signs sensing for human building interactive systems; From time-series to 2D images for building occupancy prediction using deep transfer learning; Passenger Occupancy Estimation in Vehicles: A Review of Current Methods and Research Challenges; Using Stream Data Processing for Real-Time Occupancy Detection in Smart Buildings; Multimodal sensor fusion framework for residential building occupancy detection; Efficient Occupancy Detection System Based on Neutrosophic Weighted Sensors Data Fusion; Indoor Occupancy Detection Based on Environmental Data Using CNN-XGboost Model: Experimental Validation in a Residential Building
6) The resolution of Figure 1 should be increased.
7) Inserted a figure that summarizes the taxonomy of existing occupancy detection methods and their characteristics can help help the reader understand the topic.
8) Inserting some figures that illustrate examples of occuapncy detection systems is of utmost importance. Please insert two or three flowchart of existing literature.
Author Response
Please see the attachment.
Author Response File: 
Author Response.docx
Reviewer 2 Report
This manuscript presents an intersting review of previos work regarding the use of multi-sensor (networked sensor nodes) fusion-based solutions for detection of indoor occupancy, which is focused on passive Infrared sensors. The basic description and comparison of methods is presented (based on application desired parameters for applications like health and safety and occupant comfort (i.e. beyond classically observed HVAC control). The study presenets occupancy detection criteria and introduce novel metric “sensor node deployment density” which is able to capture the strength of network-level data filtering & fusion algorithms. The survey provides a standardized insight into networked nodes' occupancy detection (includingdata collection and connectivity strategies), data fusion solution, proposed machine learning algorithms and occupancy estimation algorithms. However, the most valuable result may be that the paper highlights the gaps in past research and give certain recommendations realted to the suitability of existing methods towards a standard set of application areas.
The review paper is well conceived and written, with the main ideas, and results clearly presented. However, there are some aspects that could be improved:
- The Section 4 - Solutions to Application Mapping, mostly present main data/results in the tabelar form. Some more elaborate description and comparison/discussion could be beneficial for the reader to better grasp the background information.
- The conclusion is a bit too formal, while some of the formulations are a bit confusiong or unfinished, i.e. the last sentence which seams to be completely unsuitable. This should be revisited, and some remark on future trends/work should be highlighted in conclusion as well (these are given in disscussion section - maybe this is the reason conclusion looks like bit unessential).
- The approach to the abbreviation referencing must be improved, i.e. some but not all are given at the end of the paper, while some but around 50% of them are defined in the text. All the abbreviations should be defined in the text (at the first appereance), and the list at the end should be updated if given.
- The small number of fortmating and language errors (mostly typos) should be corrected.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
The authors have significantly improve the paper by addressing all the comments. I have no further suggestion.
