Helmet-Mounted Real-Time Toxic Gas Monitoring and Prevention System for Workers in Confined Places

Occupational health and safety hazards associated with confined places are mainly caused by exposure to toxic gases and oxygen deficiency. Lack of awareness, inappropriate monitoring, and improper evacuation methods can lead to worker fatalities. Although previous studies have attempted to develop systems to solve this issue, limited research is available on their application in confined places. In this study, a real-time helmet-mounted system was developed to monitor major toxic gases (methane (CH4), hydrogen sulfide (H2S), ammonia (NH3), and carbon monoxide (CO)), oxygen, temperature, and humidity. Workers outside and inside confined spaces receive alerts every second to immediately initiate the rescue operation in the event of a hazard. The test results of a confined environment (wastewater treatment unit) highlighted that concentrations of CH4 and H2S were predominant (13 ppm). Compared to normal atmosphere, CH4 concentration was 122- and 130-fold higher in the landfill and digestion tanks, respectively, while H2S was 36- and 19-fold higher in the primary and secondary clarifiers, respectively. The oxygen content (18.2%) and humidity (33%) were below the minimum required limits. This study will benefit future research to target appropriate toxic gas monitoring and alert workers by studying the existing issues and associated factors in confined places.


Introduction
The increasing fatality rate of construction workers in confined workplaces is becoming a growing concern. Confined places have been classified as limited or restricted for entry and exit [1], with only limited space for workers to perform drainage, cleaning, or other maintenance tasks. Workplace accidents occur for a variety of reasons, including exposure to harmful substances or environments, fire or explosion, physical risks such as (falls, slips, or trips), contact with objects or equipment, and engulfment [2,3]. A confined atmosphere is dangerous and different from the normal atmosphere as it contains various toxic gases, with low oxygen levels [4]. Toxic gas inhalation remains a major cause of injuries and deaths in confined workers.
Several previous studies have focused on confined place accidents and fatality rates. Naghavi et al. [5] found that 70.91% of fatalities in Iran (from 2006 to 2017) were due to hydrogen sulfide, nitrogen poisoning, and oxygen deficiency, with explosions, electrocutions, and poisoning outside the confined place causing the remaining fatalities. The United States National Institute for Occupational Safety and Health (NIOSH) data and Italian data reports that oxygen deficiency and toxic gas poisoning caused major deaths in confined spaces between 1985-2015 and 2001-2016, respectively [6]. According to the Occupational Safety and Health Administration (OSHA) and Mine Safety and Health Administration (MSHA) reports, asphyxiation and toxic gas inhalation were the primary reasons contributing to an accident fatality rate of 1.5. Selman et al. [2] identified that 92% of deaths in confined Sensors 2023, 23  The designed helmet-mounted safety system was tested under normal, toxic, and confined atmospheric conditions. The test results obtained were analyzed based on the standard exposure limits of the concerned authorities. Finally, the conclusions and future recommendations are provided.

Design of the Helmet-Mounted System
The effectiveness of a helmet-mounted system depends on the selection of the monitoring unit, which should be able to report real-time gas concentrations precisely and promptly to inform the worker. In this study, a helmet-mounted system was designed and developed to simultaneously measure the major toxic gases in confined places and the atmospheric conditions of the toxic gas environment. Figure 3 shows a block diagram of the proposed system. Alphasense gas sensors (such as CO/H2S, NH3-B1, CH-D3, and O2-A3), DHT11 temperature and humidity sensors, LED, and HC-05 Bluetooth module were connected to the Arduino Mega. An LED is used to provide real-time alerts to the worker inside the confined place, while the worker outside the confined place receives  The designed helmet-mounted safety system was tested under normal, toxic, and confined atmospheric conditions. The test results obtained were analyzed based on the standard exposure limits of the concerned authorities. Finally, the conclusions and future recommendations are provided.

Design of the Helmet-Mounted System
The effectiveness of a helmet-mounted system depends on the selection of the monitoring unit, which should be able to report real-time gas concentrations precisely and promptly to inform the worker. In this study, a helmet-mounted system was designed and developed to simultaneously measure the major toxic gases in confined places and the atmospheric conditions of the toxic gas environment. Figure 3 shows a block diagram of the proposed system. Alphasense gas sensors (such as CO/H2S, NH3-B1, CH-D3, and O2-A3), DHT11 temperature and humidity sensors, LED, and HC-05 Bluetooth module were connected to the Arduino Mega. An LED is used to provide real-time alerts to the worker inside the confined place, while the worker outside the confined place receives The designed helmet-mounted safety system was tested under normal, toxic, and confined atmospheric conditions. The test results obtained were analyzed based on the standard exposure limits of the concerned authorities. Finally, the conclusions and future recommendations are provided.

Design of the Helmet-Mounted System
The effectiveness of a helmet-mounted system depends on the selection of the monitoring unit, which should be able to report real-time gas concentrations precisely and promptly to inform the worker. In this study, a helmet-mounted system was designed and developed to simultaneously measure the major toxic gases in confined places and the atmospheric conditions of the toxic gas environment. Figure 3 shows a block diagram of the proposed system. Alphasense gas sensors (such as CO/H2S, NH3-B1, CH-D3, and O2-A3), DHT11 temperature and humidity sensors, LED, and HC-05 Bluetooth module were connected to the Arduino Mega. An LED is used to provide real-time alerts to the worker inside the confined place, while the worker outside the confined place receives

Real-Time Gas Monitoring
Alphasense electrochemical gas sensors were used in this study to measure the gases. Several publications have advocated using these sensors for occupational and environmental gas monitoring. Penza et al. [36] used Alphasense sensors to monitor the concentrations of nitrogen dioxide (NO2), CO, sulfur dioxide (SO2), H2S, and particulate matter (PM10), suggesting that the sensor system is cost-effective and can be used for portable air quality monitoring. These sensors have the advantage of being effectively used in smart sensor networks in smart cities. Oletic and Bilas [37] measured the concentrations of CO, SO2, and NO2 using AlphaSense sensors. Using the analog front end (AFE), the output current from the sensor is converted to voltage; thus, the gas concentration is obtained from the output voltage. This study showed that the sensors were accurate within ±0.6 ppm CO and the power consumption was lower.
Afshar-Mohajer et al. [34] developed an air quality monitoring network for measuring ozone (O3), NO2, and CO and evaluated low-cost electrochemical sensors produced by Alphasense, Ltd. (Braintree, UK). A strong linear relationship was observed between

Real-Time Gas Monitoring
Alphasense electrochemical gas sensors were used in this study to measure the gases. Several publications have advocated using these sensors for occupational and environmental gas monitoring. Penza et al. [36] used Alphasense sensors to monitor the concentrations of nitrogen dioxide (NO2), CO, sulfur dioxide (SO2), H2S, and particulate matter (PM10), suggesting that the sensor system is cost-effective and can be used for portable air quality monitoring. These sensors have the advantage of being effectively used in smart sensor networks in smart cities. Oletic and Bilas [37] measured the concentrations of CO, SO2, and NO2 using AlphaSense sensors. Using the analog front end (AFE), the output current from the sensor is converted to voltage; thus, the gas concentration is obtained from the output voltage. This study showed that the sensors were accurate within ±0.6 ppm CO and the power consumption was lower.
Afshar-Mohajer et al. [34] developed an air quality monitoring network for measuring ozone (O3), NO2, and CO and evaluated low-cost electrochemical sensors produced by Alphasense, Ltd. (Braintree, UK). A strong linear relationship was observed between

Real-Time Gas Monitoring
Alphasense electrochemical gas sensors were used in this study to measure the gases. Several publications have advocated using these sensors for occupational and environmental gas monitoring. Penza et al. [36] used Alphasense sensors to monitor the concentrations of nitrogen dioxide (NO 2 ), CO, sulfur dioxide (SO 2 ), H 2 S, and particulate matter (PM 10 ), suggesting that the sensor system is cost-effective and can be used for portable air quality monitoring. These sensors have the advantage of being effectively used in smart sensor networks in smart cities. Oletic and Bilas [37] measured the concentrations of CO, SO 2 , and NO 2 using AlphaSense sensors. Using the analog front end (AFE), the output current from the sensor is converted to voltage; thus, the gas concentration is obtained from the output voltage. This study showed that the sensors were accurate within ±0.6 ppm CO and the power consumption was lower.
Afshar-Mohajer et al. [34] developed an air quality monitoring network for measuring ozone (O 3 ), NO 2 , and CO and evaluated low-cost electrochemical sensors produced by Alphasense, Ltd. (Braintree, UK). A strong linear relationship was observed between the output voltages and gas concentrations (i.e., (R 2 > 0.98). CO sensors were considered suitable for occupational and environmental CO monitoring because of their field-ready calibration capabilities. Marinov et al. [35] conducted a study that monitored NO 2 , O 3 , CO, and SO 2 levels in the atmosphere using four-electrode amperometric gas sensors manufactured by Alphasense, with a four-sensor AFE board used for analog-to-digital conversion. These sensors were found to be applicable for low-power, low-cost, and fast-response air quality monitoring. The developed sensor interface was tested, and its accuracy was compared with that of the multichannel data acquisition (DAQ) system.
Sharma et al. [38] developed a low-cost sensor system by utilizing Alphasense electrochemical gas sensors to monitor NO 2 , O 3 , and PM 2.5 levels in urban air. The findings of this study highlighted that the raw sensor data showed a determination coefficient (R 2 ) of >0.80. The power consumption of the sensor system was under 0.5 W, leading to the authors recommending a sensor network for wireless systems in the field of air quality monitoring. Arroyo et al. [39] measured NO 2 , CO, O 3 , PM 2.5 , and PM 10 concentrations in an outdoor environment. The R 2 values obtained for the sensor data were in the range of 0.83-0.95, suggesting that the sensor system is suitable for portable air quality monitoring, with higher accuracy and noise reduction achievable by carefully designing the sensor interface.
In this study, a printed circuit board (PCB) was designed to combine gas sensors into a single unit. The output current from the Alphasense gas sensors was converted by employing an LMP91000 AFE, which helps signal communication between the sensor and microcontroller and converts the output cell current proportional to the output voltage. In addition, it supports programming with multiple electrochemical gas sensors and has a gas sensitivity range of 0.5-9500 nA/ppm, operating temperature between −40 • C to 85 • C, and voltage supply of 2.7-5.25 V. LMP91000 consumes a total current of 10µA and helps in noise reduction [40]. The developed sensor unit and connections of the PCB with the Arduino Mega are shown in Figure 5. The numbers 1 to 8 indicate the PINs of PCB for each header (i.e., H1, H2, H3, H4, H5, and H6).
Sensors 2023, 23, x FOR PEER REVIEW 6 of 16 the output voltages and gas concentrations (i.e., (R 2 > 0.98). CO sensors were considered suitable for occupational and environmental CO monitoring because of their field-ready calibration capabilities. Marinov et al. [35] conducted a study that monitored NO2, O3, CO, and SO2 levels in the atmosphere using four-electrode amperometric gas sensors manufactured by Alphasense, with a four-sensor AFE board used for analog-to-digital conversion. These sensors were found to be applicable for low-power, low-cost, and fast-response air quality monitoring. The developed sensor interface was tested, and its accuracy was compared with that of the multichannel data acquisition (DAQ) system. Sharma et al. [38] developed a low-cost sensor system by utilizing Alphasense electrochemical gas sensors to monitor NO2, O3, and PM2.5 levels in urban air. The findings of this study highlighted that the raw sensor data showed a determination coefficient (R 2 ) of >0.80. The power consumption of the sensor system was under 0.5 W, leading to the authors recommending a sensor network for wireless systems in the field of air quality monitoring. Arroyo et al. [39] measured NO2, CO, O3, PM2.5, and PM10 concentrations in an outdoor environment. The R 2 values obtained for the sensor data were in the range of 0.83-0.95, suggesting that the sensor system is suitable for portable air quality monitoring, with higher accuracy and noise reduction achievable by carefully designing the sensor interface.
In this study, a printed circuit board (PCB) was designed to combine gas sensors into a single unit. The output current from the Alphasense gas sensors was converted by employing an LMP91000 AFE, which helps signal communication between the sensor and microcontroller and converts the output cell current proportional to the output voltage. In addition, it supports programming with multiple electrochemical gas sensors and has a gas sensitivity range of 0.5-9500 nA/ppm, operating temperature between −40°C to 85 °C, and voltage supply of 2.7-5.25 V. LMP91000 consumes a total current of 10μA and helps in noise reduction [40]. The developed sensor unit and connections of the PCB with the Arduino Mega are shown in Figure 5. The numbers 1 to 8 indicate the PINs of PCB for each header (i.e., H1, H2, H3, H4, H5, and H6). The sensor unit was maintained inside a 3D mold and placed on the front side of the helmet. Workers were directly exposed to hazardous gases near their breathing zones. The sensor unit was maintained inside a 3D mold and placed on the front side of the helmet. Workers were directly exposed to hazardous gases near their breathing zones. Regarding the safe level of breathing, sensors should be placed near the typical breathing zone, i.e., 4-6 feet from the ground [41]. Hence, the gas sensor unit was positioned at the front of the helmet to continuously and accurately measure the gas exposure concentrations.

Temperature and Humidity Monitoring
The temperature and humidity of the confined environment are essential factors to be considered in the monitoring unit. The DHT11 temperature and humidity sensor was chosen for this purpose, with previous publications highly recommending using these sensors. The accuracy of the temperature and humidity measurements were ±2 • C and ±5% RH, respectively [42]. This sensor, comprising an 8-bit microcontroller, provides a calibrated digital output, precise temperature and humidity values, and responds quickly to exposure conditions. The DHT11 sensor has advantages such as low cost, simple interface, compactness, reliability, and stability [43][44][45]. In addition, Novelan and Amin [46], Saputro and Yantidewi [47], and Singh [48] utilized and suggested sensors for measuring temperature and humidity.
Similar to the gas sensor unit, the DHT11 sensor was mounted on the front of the helmet near it because the atmospheric parameters experienced by workers must be measured simultaneously. Figure 6 shows the connections between the sensor and microprocessor (MCU). The PINs 1, 2, 3, and 4 indicate power input (VCC), data line, no connection (NC), and ground (Gnd), respectively. Regarding the safe level of breathing, sensors should be placed near the typical breathing zone, i.e., 4-6 feet from the ground [41]. Hence, the gas sensor unit was positioned at the front of the helmet to continuously and accurately measure the gas exposure concentrations.

Temperature and Humidity Monitoring
The temperature and humidity of the confined environment are essential factors to be considered in the monitoring unit. The DHT11 temperature and humidity sensor was chosen for this purpose, with previous publications highly recommending using these sensors. The accuracy of the temperature and humidity measurements were ±2 °C and ±5% RH, respectively [42]. This sensor, comprising an 8-bit microcontroller, provides a calibrated digital output, precise temperature and humidity values, and responds quickly to exposure conditions. The DHT11 sensor has advantages such as low cost, simple interface, compactness, reliability, and stability [43][44][45]. In addition, Novelan and Amin [46], Saputro and Yantidewi [47], and Singh [48] utilized and suggested sensors for measuring temperature and humidity.
Similar to the gas sensor unit, the DHT11 sensor was mounted on the front of the helmet near it because the atmospheric parameters experienced by workers must be measured simultaneously. Figure 6 shows the connections between the sensor and microprocessor (MCU). The PINs 1,2,3, and 4 indicate power input (VCC), data line, no connection (NC), and ground (Gnd), respectively.

Wireless Communication
The exposure conditions and corresponding alert information must be relayed to the worker outside the confined place to avoid danger. For this purpose, the HC-05 Bluetooth module is utilized, which works on Serial Port Protocol (SPP) [49]. This module has a 2.4 GHz built-in antenna and operates between −25 to 75 °C. In addition, its compact size, easy connection with Bluetooth-enabled devices, better operating life, and low power consumption make it suitable for wireless communication.
Yadav and Vohra [32] analyzed the performance of an HC-05 Bluetooth module with an ESP8266 Wi-Fi module, with HC-05 exhibiting better battery utilization and efficiency than ESP8266. The Wi-Fi module used more current than HC-05, resulting in a shorter operating life. The modulation and transmit powers of HC-05 are 3 Mbps and +4 dBm RF, respectively [50]. The HC-05 module is cost-effective, sending signals quickly and with better quality [51]. The connection diagram of HC-05 with the Arduino Mega is illustrated in Figure 7. Power input or VCC, GND, transmit data or TXD, and receive data or RXD are represented by red, black, green, and yellow cables, respectively.

Wireless Communication
The exposure conditions and corresponding alert information must be relayed to the worker outside the confined place to avoid danger. For this purpose, the HC-05 Bluetooth module is utilized, which works on Serial Port Protocol (SPP) [49]. This module has a 2.4 GHz built-in antenna and operates between −25 to 75 • C. In addition, its compact size, easy connection with Bluetooth-enabled devices, better operating life, and low power consumption make it suitable for wireless communication.
Yadav and Vohra [32] analyzed the performance of an HC-05 Bluetooth module with an ESP8266 Wi-Fi module, with HC-05 exhibiting better battery utilization and efficiency than ESP8266. The Wi-Fi module used more current than HC-05, resulting in a shorter operating life. The modulation and transmit powers of HC-05 are 3 Mbps and +4 dBm RF, respectively [50]. The HC-05 module is cost-effective, sending signals quickly and with better quality [51]. The connection diagram of HC-05 with the Arduino Mega is illustrated in Figure 7. Power input or VCC, GND, transmit data or TXD, and receive data or RXD are represented by red, black, green, and yellow cables, respectively.

Alert System and Decision Criteria
An LED was provided at the front of the safety helmet to alert the worker inside th confined area. The decision criteria for the toxic gas exposure prevention system, base on the gas levels, are shown in Figure 8. The threshold limits of the gases are listed i Table 1 [52][53][54][55]. The short-term exposure limit (STEL) is the 15 min exposure that shoul not be increased at any time during the working day. The time-weighted average (TWA refers to the maximum gas concentration for 8 h over a five-working-day week. Immed ately Dangerous to Life or Health (IDLH) exposure standard is the peak concentration a which the exposure can be fatal if the worker is not equipped with sufficient respirator protection units. In the current study, STEL was considered for the alert as real-time con centrations measured cannot be exceeded at that particular time. In the case of long-term monitoring, TWA can be utilized because it quantifies the mean exposure to the worker during a time period. For CH4, the maximum threshold limit is 1000 ppm during an 8 working shift, as set by NIOSH. The alert system is activated when the toxic gas levels ar higher and the oxygen concentration is lower than the recommended limits. If the level are safe according to the standards, the alert system will be turned off, and the proces will continue.
The developed helmet-mounted system is illustrated in Figure 9. The sensor system is portable, connected to a 9 V power supply and the Arduino can be powered for abou 10 h to obtain the sensor data continuously. The data are transferred to the mobile devic every second, and an alert is provided simultaneously. Therefore, rescue operations ca be performed quickly to save workers' lives without further hazardous gas exposure.

Alert System and Decision Criteria
An LED was provided at the front of the safety helmet to alert the worker inside the confined area. The decision criteria for the toxic gas exposure prevention system, based on the gas levels, are shown in Figure 8. The threshold limits of the gases are listed in Table 1 [52][53][54][55]. The short-term exposure limit (STEL) is the 15 min exposure that should not be increased at any time during the working day. The time-weighted average (TWA) refers to the maximum gas concentration for 8 h over a five-working-day week. Immediately Dangerous to Life or Health (IDLH) exposure standard is the peak concentration at which the exposure can be fatal if the worker is not equipped with sufficient respiratory protection units. In the current study, STEL was considered for the alert as real-time concentrations measured cannot be exceeded at that particular time. In the case of long-term monitoring, TWA can be utilized because it quantifies the mean exposure to the workers during a time period. For CH 4 , the maximum threshold limit is 1000 ppm during an 8 h working shift, as set by NIOSH. The alert system is activated when the toxic gas levels are higher and the oxygen concentration is lower than the recommended limits. If the levels are safe according to the standards, the alert system will be turned off, and the process will continue. The developed helmet-mounted system is illustrated in Figure 9. The sensor system is portable, connected to a 9 V power supply and the Arduino can be powered for about 10 h to obtain the sensor data continuously. The data are transferred to the mobile device every second, and an alert is provided simultaneously. Therefore, rescue operations can be performed quickly to save workers' lives without further hazardous gas exposure.

Testing of the Helmet-Mounted System
The helmet-mounted system was tested in normal and confined environments to compare the fluctuations in the sensor readings, with a wastewater treatment plant used for testing. The Jangrim wastewater treatment plant in Busan is a major municipal and industrial wastewater treatment plant in South Korea that collects and treats approximately 450 tons of wastewater. Figure 10 shows the units in the wastewater treatment plant where measurements were conducted-landfill area, primary sedimentation tank, secondary clarifier, and sludge digestion tank. Monitoring was performed for 10 min in both normal and toxic gas environments with a time interval of one second. If the atmospheric condition is dangerous, "Alert ON/danger" is sent to the mobile device, and safe conditions are displayed as "Alert OFF/Safe". Therefore, workers outside the confined space can properly assess the situation accordingly.

Testing of the Helmet-Mounted System
The helmet-mounted system was tested in normal and confined environments to compare the fluctuations in the sensor readings, with a wastewater treatment plant used for testing. The Jangrim wastewater treatment plant in Busan is a major municipal and industrial wastewater treatment plant in South Korea that collects and treats approximately 450 tons of wastewater. Figure 10 shows the units in the wastewater treatment plant where measurements were conducted-landfill area, primary sedimentation tank, secondary clarifier, and sludge digestion tank. Monitoring was performed for 10 min in both normal and toxic gas environments with a time interval of one second. If the atmospheric condition is dangerous, "Alert ON/danger" is sent to the mobile device, and safe conditions are displayed as "Alert OFF/Safe". Therefore, workers outside the confined space can properly assess the situation accordingly.

Testing of the Helmet-Mounted System
The helmet-mounted system was tested in normal and confined environments to compare the fluctuations in the sensor readings, with a wastewater treatment plant used for testing. The Jangrim wastewater treatment plant in Busan is a major municipal and industrial wastewater treatment plant in South Korea that collects and treats approximately 450 tons of wastewater. Figure 10 shows the units in the wastewater treatment plant where measurements were conducted-landfill area, primary sedimentation tank, secondary clarifier, and sludge digestion tank. Monitoring was performed for 10 min in both normal and toxic gas environments with a time interval of one second. If the atmospheric condition is dangerous, "Alert ON/danger" is sent to the mobile device, and safe conditions are displayed as "Alert OFF/Safe". Therefore, workers outside the confined space can properly assess the situation accordingly.   Figure 11 represent the data for the gases obtained and the exposure conditions in the normal atmosphere. The average and maximum concentrations recorded were 0.28 ppm and 0.36 ppm for H2S, 0.29 ppm and 0.35 ppm for CO, 0.19 ppm and 0.26 ppm for NH3, and 0.07 and 0.1 ppm for CH4 (Figure 11a). All toxic gas levels were under 1 ppm, considered safe according to the threshold limits. The mean humidity and temperature were 41% and 25.2 °C, with maximum levels of 42% and 25.6 °C , respectively. A humidity range of 40-70% is considered less hazardous to workers in confined places [56]. The optimal temperature for the working conditions was below 30 °C . Workplace temperatures of ≥30 °C can cause discomfort and heat stress. Hence, the temperature and humidity in a normal environment were found to be within these limits. An average oxygen concentration of 20.15% was obtained in a normal environment, higher than the recommended minimum level of 19.5% (Figure 11b).    Figure 11 represent the data for the gases obtained and the exposure conditions in the normal atmosphere. The average and maximum concentrations recorded were 0.28 ppm and 0.36 ppm for H 2 S, 0.29 ppm and 0.35 ppm for CO, 0.19 ppm and 0.26 ppm for NH 3 , and 0.07 and 0.1 ppm for CH 4 (Figure 11a). All toxic gas levels were under 1 ppm, considered safe according to the threshold limits. The mean humidity and temperature were 41% and 25.2 • C, with maximum levels of 42% and 25.6 • C, respectively. A humidity range of 40-70% is considered less hazardous to workers in confined places [56]. The optimal temperature for the working conditions was below 30 • C. Workplace temperatures of ≥30 • C can cause discomfort and heat stress. Hence, the temperature and humidity in a normal environment were found to be within these limits. An average oxygen concentration of 20.15% was obtained in a normal environment, higher than the recommended minimum level of 19.5% (Figure 11b).  The recorded hazardous gas concentrations in various units of the wastewater treatment plant are shown in Figure 12. An increasing trend of gas concentrations was obtained, which reached a stable value after some time. The toxic gas data were initially inclined, with measurements taken for 10 min to obtain stable values. In a wastewater treatment plant, wastewater first undergoes a preliminary treatment where large objects, such as plastic, rags, and rubbish, are removed using special filter screens. These wastes were collected in a landfill area. A high concentration of CH4 was generated in the landfill area because of the natural bacterial decomposition of organic matter. The mean and maximum levels of CH4 were 8.21 ppm and 12.22 ppm, respectively. Followed by CH4, H2S showed mean and maximum exposure concentrations of 4.15 ppm and 5.72, while NH3 concentrations were 3.53 ppm and 4.62 ppm, respectively (Figure 12a). The recorded hazardous gas concentrations in various units of the wastewater treatment plant are shown in Figure 12. An increasing trend of gas concentrations was obtained, which reached a stable value after some time. The toxic gas data were initially inclined, with measurements taken for 10 min to obtain stable values. In a wastewater treatment plant, wastewater first undergoes a preliminary treatment where large objects, such as plastic, rags, and rubbish, are removed using special filter screens. These wastes were collected in a landfill area. A high concentration of CH 4 was generated in the landfill area because of the natural bacterial decomposition of organic matter. The mean and maximum levels of The recorded hazardous gas concentrations in various units of the wastewater treatment plant are shown in Figure 12. An increasing trend of gas concentrations was obtained, which reached a stable value after some time. The toxic gas data were initially inclined, with measurements taken for 10 min to obtain stable values. In a wastewater treatment plant, wastewater first undergoes a preliminary treatment where large objects, such as plastic, rags, and rubbish, are removed using special filter screens. These wastes were collected in a landfill area. A high concentration of CH4 was generated in the landfill area because of the natural bacterial decomposition of organic matter. The mean and maximum levels of CH4 were 8.  After filtering, the wastewater is passed into the primary sedimentation tank, settling tank, or clarifier to remove the particles from the wastewater, and sludge is settled at the bottom of the tank. After primary sedimentation in the sedimentation tank, the floating sludge from the wastewater was collected in the secondary clarifier, and the total sludge is taken to the sludge treatment area. H 2 S exhibited the highest concentration in both units (Figure 12b,c). The average mean and maximum exposure values were 9.27 ppm and 13 ppm in the primary sedimentation tank and 4.22 ppm and 7 ppm in the secondary clarifier, respectively. The second-highest average and maximum levels were for CH 4, with values of 8.15 ppm and 10.5 ppm for the primary clarifier and 2.53 ppm and 4.3 ppm for the secondary clarifier. NH 3 showed a maximum concentration of 5.25 ppm and 2.08 ppm for primary and secondary clarifiers, respectively (Figure 12b,c).
Sludge treatment was performed in digestion tanks where anaerobic digestion was performed by bacteria. CH 4 was the primary gas generated during the anaerobic digestion process, similar to the results obtained for the landfill area. The mean and maximum concentrations of CH 4 were 9.47 ppm and 13 ppm, respectively. For NH 3 and H 2 S, the mean concentrations recorded were 2.5 ppm and 1.97 pm, and maximum levels were 3.36 ppm and 2.42 ppm, respectively (Figure 12d). CO concentrations were the lowest in all four monitored units. The maximum exposure values were 2.24 ppm, 2.52 ppm, 1, and 1.25 ppm in the landfill, primary sedimentation tank, secondary clarifiers, and digestion tanks, respectively. However, all the gas exposures were within the threshold limits set by the concerned authorities. Figure 13 shows the atmospheric conditions of the wastewater treatment and toxic gas environment. Mean values of 33% and 19.17 • C were determined for humidity and temperature, respectively, less than that in normal atmospheric conditions while the humidity was lower than the recommended value. The obtained mean oxygen level of 18.2% was lower than the minimum required level of 19.5%. After filtering, the wastewater is passed into the primary sedimentation tank, settling tank, or clarifier to remove the particles from the wastewater, and sludge is settled at the bottom of the tank. After primary sedimentation in the sedimentation tank, the floating sludge from the wastewater was collected in the secondary clarifier, and the total sludge is taken to the sludge treatment area. H2S exhibited the highest concentration in both units (Figure 12b,c). The average mean and maximum exposure values were 9.27 ppm and 13 ppm in the primary sedimentation tank and 4.22 ppm and 7 ppm in the secondary clarifier, respectively. The second-highest average and maximum levels were for CH4, with values of 8.15 ppm and 10.5 ppm for the primary clarifier and 2.53 ppm and 4.3 ppm for the secondary clarifier. NH3 showed a maximum concentration of 5.25 ppm and 2.08 ppm for primary and secondary clarifiers, respectively (Figure 12b,c).
Sludge treatment was performed in digestion tanks where anaerobic digestion was performed by bacteria. CH4 was the primary gas generated during the anaerobic digestion process, similar to the results obtained for the landfill area. The mean and maximum concentrations of CH4 were 9.47 ppm and 13 ppm, respectively. For NH3 and H2S, the mean concentrations recorded were 2.5 ppm and 1.97 pm, and maximum levels were 3.36 ppm and 2.42 ppm, respectively (Figure 12d). CO concentrations were the lowest in all four monitored units. The maximum exposure values were 2.24 ppm, 2.52 ppm, 1, and 1.25 ppm in the landfill, primary sedimentation tank, secondary clarifiers, and digestion tanks, respectively. However, all the gas exposures were within the threshold limits set by the concerned authorities. Figure 13 shows the atmospheric conditions of the wastewater treatment and toxic gas environment. Mean values of 33% and 19.17 °C were determined for humidity and temperature, respectively, less than that in normal atmospheric conditions while the humidity was lower than the recommended value. The obtained mean oxygen level of 18.2% was lower than the minimum required level of 19.5%.

Discussion
The construction industry has been identified as a major source of workplace accidents and hazardous exposure fatalities. Physical hazards, toxic atmospheres, and engulfment can emerge as occupational health and safety risks. Hazardous gases are deadly and harmful when they enter confined atmospheres. Confined spaces are fully or partially enclosed areas in which only a limited number of workers can enter to perform assigned activities. Toxic gas accumulation, oxygen deficiency, poor ventilation, mechanical

Discussion
The construction industry has been identified as a major source of workplace accidents and hazardous exposure fatalities. Physical hazards, toxic atmospheres, and engulfment can emerge as occupational health and safety risks. Hazardous gases are deadly and harmful when they enter confined atmospheres. Confined spaces are fully or partially enclosed areas in which only a limited number of workers can enter to perform assigned activities. Toxic gas accumulation, oxygen deficiency, poor ventilation, mechanical equipment usage, and flammable atmosphere can make confined atmospheres more dangerous than normal environments. Short-term and long-term exposure to hazardous gases creates various health issues, such as headaches, vomiting, unconsciousness, sudden collapses, and even death. Previous studies have highlighted that the hazards posed by toxic gases lead to the majority of confined place fatalities; thus, more attention is needed to properly mitigate this issue.
In this study, we developed a helmet-mounted toxic gas monitoring and prevention system that can measure the primary toxic gases in confined places such as H 2 S, NH 3 , CH 4 , and CO using Alphasense gas sensors; these sensors are suitable for low-cost, accurate, stable, fast response, low power, and portable occupational and environmental monitoring [34][35][36][37][38][39]. Oxygen level in air is also an important parameter to be considered in confined spaces. Alphasense CO/H2S D2, CH4-D3, NH3-B1, and O2-A3 sensors were employed for measuring concentrations of these gases, using a sensor unit designed and developed by the authors. Other exposure conditions, such as humidity and temperature, were also assessed using the DHT11 sensor. The monitoring unit was positioned at the front side of the helmet for measuring concentration levels near worker breathing zones every second. The toxic gas prevention system comprises an HC-05 Bluetooth module to provide real-time notifications to the worker or supervisor outside the confined location, while an LED alert is provided to the worker inside the confined location based on the developed decision criteria. As alerts are given within each second, rescue operations can be performed smoothly and quickly with proper usage of rescue equipment.
Testing was conducted in four units of the wastewater treatment plant: landfill, primary sedimentation tank, secondary clarifier, and digestion tank. A higher concentration of CH 4 was recorded in the landfill area and sludge digestion tank, while the H 2 S content was higher in the primary and secondary clarifiers. The gases were mainly produced by the decomposition and anaerobic digestion processes. However, all gas levels were safe and below the threshold limits, with CO concentration the lowest in all the units. The maximum concentrations of CH 4 in the landfill and digestion tanks were 122 and 130 times higher than the normal environment, respectively. In contrast, H 2 S levels were 36 and 19 times higher in the primary sedimentation tank and secondary clarifier, respectively. The oxygen level and humidity in the wastewater treatment units (18.2% and 33%, respectively) were lower than the recommended minimum limit. Thus, these results indicated fluctuations in gas concentrations and other exposure conditions of confined atmospheres compared with normal environments.
The current study included safety monitoring of toxic gas exposure for workers involved in confined spaces. The findings of this study will allow future researchers to broaden the safety monitoring system to a health and safety monitoring system by integrating workers' health conditions such as pulse rate, blood oxygen level, body temperature, and EEG or brainwave analysis. The effects of humidity and temperature on sensor readings, comparison of sensor data with bench march devices, short-and long-term exposure assessments, sensor dynamics analysis, and displaying system status can enhance the performance and efficiency of the monitoring system. Combining toxic gas monitoring and the measurement of other major pollutants in the construction industry, such as particulate matter, can mitigate occupational health impacts at construction sites.
Author Contributions: Conceptualization, methodology, software, validation, formal analysis, writing-original draft preparation, writing-review and editing, visualization, J.P.P.R.; supervision, project administration, funding acquisition, J.-h.C. All authors have read and agreed to the published version of the manuscript.