This study presents an IoT-based real-time indoor air quality monitoring system designed for laboratory classrooms to provide students with optimal working and learning conditions. The system uses a Raspberry Pi as a central hub for data collection, processing, and visualization, paired with sensors such as the MQ135 and DHT22 to monitor key environmental parameters, including ammonia (NH3), CO2, benzene, smoke, temperature, and humidity, in parts per million (ppm). Sensor data is collected and stored in a PostgreSQL database, and real-time visualization is performed using a Django web application. The results show elevated levels of contaminants during long soldering sessions, highlighting the need for effective ventilation strategies. The integration of the Raspberry Pi system improves the accuracy and responsiveness of air quality monitoring, providing a scalable and cost-effective solution for maintaining a safe indoor environment. Future work aims to optimize data processing algorithms and integrate advanced analytics to predict and proactively address air quality issues.
Most individuals spend their substantial amount of time in indoor activities, therefore maintaining
indoor air quality is critical. Whether working in enclosed offices or open places with a large number
of employees, air quality must be monitored and controlled to improve working conditions and
employee well-being. Maintaining air quality is critical in modern educational laboratories,
particularly those with soldering stations, to ensure a safe and optimal working and learning
environment. In academic institutions, for the practical purpose of teaching electronics, electrical
circuits are assembled, where soldering stations are often used. However, long-term use of soldering
equipment can lead to the accumulation of hazardous materials in the air, which can be harmful to
the health of teachers and students. In addition, such environmental elements as humidity and
temperature [
There are many sources of indoor air pollution, such as heating appliances, tobacco products,
building materials, central heating and cooling systems, humidifiers, excess moisture, and outdoor
sources [
The authors [13] indicate that real-time monitoring of classroom CO2 levels can be used as a proxy for the risk of SARS-CoV-2 transmission. They implemented a customized ventilation protocol with real-time CO2 monitoring, improving CO2 levels in all classrooms where teachers followed it. The study [14] examined indoor environmental quality, including indoor air quality, ventilation requirements, and health impacts assessed using the Cancer Hazard Index and Risk of Cancer in a naturally ventilated school. The study examined the relationship between ventilation, CO2, and particulate matter (PM) levels. It assessed the potential health hazards of pollutants to students using the US Environmental Protection Agency's Cancer Hazard Index and Risk of Cancer. The paper [15] presents an IoT-based system for continuous monitoring and assessment of indoor air quality (IAQ) in an educational building, which includes collecting real-time measurements of CO2, CO, and PM2.5 parameters, transmitting the data to a cloud platform and developing a deep learning model to predict indoor environmental conditions.
Recent research on the development of IoT-based indoor air quality monitoring and control systems uses low-cost sensors to measure pollutants such as CO, CO2, and particulate matter, as well as readily available microcontrollers and communication modules to process and transmit data. Such projects include alarm systems and ventilation activation based on pollutant levels [16]. The authors [17] propose a system that collects, processes and transmits air quality indices to servers for storage and visualization. IoT platforms with low-cost sensors have shown good potential in improving indoor air quality management, and regular replacement of sensors is recommended for reliability. These systems can integrate data-driven algorithms for IAQ prediction and ventilation control, balancing energy efficiency with air quality improvement [18].
The rest of the article is organized as: Section 2 presents methods and materials. Section 3 gives an overview of experimental results, and Section 4 concludes the entire article and provides future work.
The study describes a system that is specifically intended for laboratory classrooms and uses a Raspberry Pi as a central hub for data collecting, processing, and display. This system uses sensors such as the MQ135 and DHT22 to monitor vital environmental parameters like ammonia (NH3), CO2, benzene, smoke, temperature, and humidity instantaneously. The data is saved in a PostgreSQL database and shown via a Django web application, providing an adaptable and cost-effective approach for ensuring a safe indoor environment. The findings of the study reveal that the monitoring system successfully detected high levels of pollutants, particularly during long soldering sessions, emphasizing the importance of proper ventilation in laboratory conditions. This contribution focuses on the system's practical applicability in improving air quality while also providing optimal working and learning environments for students.
This study was conducted in the university's Electronics laboratory room, which has a total area of around 36 m2. The laboratory room has radiators for winter heating and one large opening window (see Figure 1).
Developing a low-cost IoT-based system for real-time air quality monitoring. Collecting experimental data using sensors such as CO2, temperature, and humidity. Data processing involves organizing and cleaning the data and preparing it for further analysis.
Installing the monitoring system in a university laboratory to ensure a safe and optimal learning environment for students and teachers.
The developed system for monitoring air quality in a laboratory room includes two main components. The first part of the system consists of sensors installed in the laboratory that measure air parameters such as gas concentrations. Data from the sensors is collected using an Arduino Uno microcontroller, which reads the values and transmits them to the next part of the system. The second part is a web server running on a Raspberry Pi that collects, processes, stores, and displays data. Raspberry Pi connects to the microcontroller via a serial interface or over a network and receives sensor data. The web server allows the storage of data in a local database, displays data in real-time via a web interface, sets up alerts for exceeding permissible values of pollutants or other parameters, and provides access to historical data for analyzing changes in air quality over time. The system allows teachers and students to monitor air quality in real-time, ensuring safe working and learning conditions in the laboratory room. With the help of a web server on the Raspberry Pi, it is possible to collect and display data from sensors and control connected IoT devices in real-time. The system architecture is shown in Figure 2.
Sensor data collection and transmission are critical in Raspberry Pi-based systems since they allow for real-time environmental monitoring and analysis. The central node here is the Raspberry Pi, effectively processing and storing data acquired from numerous sensors, allowing for timely responses to environmental changes. This prototype configuration enables fast and efficient identification of hazardous pollutants or deviations from optimal conditions. It also includes automatic notifications to ensure safety. In addition, remote access and real-time data display are possible with a connection to a web server implemented on Raspberry Pi. The prototype is scalable, such a solution is suitable for various conditions, not only laboratory, but also industrial. The data from the Sd sensors is collected as follows:
n Sd=( ∑ Gas Concentrationi+Temperaturei+ Humidityi , i=1 (1) Data transmission process can be determined as follows:
n Transmission= ArduinoUno ( ∑i=1 Sdi )⟶ Serial Interface / N ⟶ where, N denotes the network, n is the number of sensors, Gas Concentrationi that includes measurements of ammonia (NH3), CO2, benzene, and smoke, Temperaturei and Humidityi are the environmental parameters that are measured.
The Arduino Uno reads and sends the data gathered to the Raspberry Pi via a serial interface or a network connection. The data processing on a Raspberry Pi is an essential component of IoT and embedded systems, which offers a capable and inexpensive platform for performing a variety of computing tasks. The adaptable architecture of the Raspberry Pi allows it to collect and process data from several sensors at the same time, which helps for real-time applications. Once sensor data is sent to the Raspberry Pi, it can execute a variety of processing tasks, ranging from simple filtering and aggregation to more advanced analytics, depending on the application needs. The data after processing is stored locally on the Raspberry Pi, and can be transferred to cloud services or viewed in real time using a connected display or web interface. Thus, data process Pd can be calculated as follows:
Pd=W s+( Raspberry Pi ( Sd )) , (3) where, W s is the web server.
Real-time visualization on the Raspberry Pi with IoT support improves the device's operation and user experience. Getting quick feedback on data and visualization is the advantage of this work. Users can monitor environmental conditions, system performance, and other important parameters as they occur. This fast access to data guarantees that anomalies, such as unsafe amounts of pollution in an air monitoring system can be discovered and handled immediately that prevents possible hazards. Furthermore, real-time visualization on a Raspberry Pi allows for better decision-making because users can observe the direct impact of environmental or system changes on the data being watched. Thus, real-time visualization V rt can be determined as follows:
V rt=W I ( Pd )+ Dbs+ H , (4) where W I denotes web interface, H is the Hist, and Dbs is the database storage The sensors used in the architecture are the MQ135 gas sensor and the DHT22 temperature and humidity sensor (see Figure 3). The MQ135 is a low-cost indoor electrochemical gas sensor, which measures ammonia, nitrogen, oxygen, alcohols, aromatics, sulfide, and smoke together as an IAQ. The sensor must warm up for at least 24-48 hours to obtain stable gas readings. The control is carried out by an Arduino Uno board equipped with an ATmega328P microcontroller. This microcontroller operates at a clock rate of up to 16 MHz and has 32 KB of flash memory for programs, 2 KB of RAM (SRAM), and 1 KB of non-volatile memory (EEPROM). For the MQ135 sensor, we apply data cleaning since the values can fluctuate due to noise or electromagnetic interference, so a moving average filter was used.
The environmental data collection interval was one hour. Table 1 shows the main characteristics of the sensors.
The article uses a hardware infrastructure with a Raspberry Pi 3 single-board computer hosting the web server. This third-generation model, released in February 2016, features a 1.2 GHz 64-bit quad-core ARM Cortex-A53 processor, 1 GB of RAM, and integrated 802.11n Wi-Fi and Bluetooth 4.1 modules, making it a powerful and versatile choice for various applications. The Raspberry Pi 3 web server receives data from an Arduino Uno microcontroller via a serial interface. This data collected from the sensors is stored in a PostgreSQL database. PostgreSQL, also known as Postgres, is a powerful, reliable, and flexible open-source relational database management system. Users have access to real-time sensor readings, graphical data representations, and analysis tools through a Django web interface. The web server running on the Raspberry Pi 3 is secured by scanning and monitoring important system parameters, and measures are taken to protect against DDoS attacks and SQL injection threats, including Nmap and the OpenAI API. Figure 4 shows a prototype of a lowcost monitoring system, and Figure 5 provides a system flow chart.
The process starts with collecting data from the sensors connected to the Arduino board. After collecting the data, a verification process is carried out. After verification, the Arduino board captures data from the MQ135 and the DHT22 sensor. The captured data is then transmitted to the Raspberry Pi, where it is confirmed to have been received. The data is then saved to a PostgreSQL database on the Raspberry Pi web server, after which a successful save check is performed. If the data is saved correctly, it is then moved on to analysis. During the study, the data is checked for vulnerabilities, and if any are found, protective measures are taken. If no vulnerabilities are found, the data is displayed on the website.
Air quality data can be effectively analyzed and used for timely decision-making, ensuring a reliable and safe air quality monitoring system, the whole procedure is shown in Algorithm 1.
S: system; D: sensor data; V: validity of data; R: data received status; DB: PostgreSQL database; A: analyzed data;
Vuln: data vulnerability status. o o o o o o o 2. Start 3. Sensor Data Collection:
o D←Collect sensor data 4. Is Data Valid?: o If V = True then Send Data to Arduino Board: D←Send to Arduino Send Data to Raspberry Pi: D←Send to Raspberry Pi 5. Data Received?: o If R = True then Store Data in PostgreSQL Database: DB←Store(D) 6. Is Data Stored?: o If DB is successfully stored then Analyze Data: A←Analyze(DB) 7. Is Data Vulnerable?: o If Vuln = True then Apply Protection Methods: Apply protection methods(A) 8. End o
Display(A)
Algorithm 1 shows the process of collecting and analyzing air quality data. Step 1 explains the initialization of variables. Steps 2–4 cover collecting and validating sensor data, where valid data is sent to both the Arduino and Raspberry Pi. Steps 5–6 describe storing the data in a PostgreSQL database and further analyzing it. Step 7 focuses on validating the data for vulnerabilities and applying protection if necessary. Step 8 displays the analyzed data on the website if no vulnerabilities are found.
Air quality: This main project directory contains settings, URLs, and other configurations. manage.py: This command-line utility allows interaction with the Django project. __init__.py: This empty file signifies that this directory should be treated as a Python package. settings.py: This file contains configuration settings for the Django project. urls.py: This file contains URL declarations for the Django project. wsgi.py: This file contains the configuration for the WSGI (Web Server Gateway Interface) used to serve the project in production. asgi.py: This file contains the configuration for the ASGI (Asynchronous Server Gateway Interface) used for async-capable web servers.
As a result, the project's structure encompasses a complete system consisting of sensors, a microcontroller, a web server, and a user interface, delivering dependable monitoring and analysis of indoor air quality. Users must have the same netmask and default gateway and know the login and password information to connect to the database. All information is secured using MD5 hash.
The web dashboard, designed using the Django framework, displays environmental data, including temperature, humidity, air quality, and CO2 levels. It uses dynamic data visualization with Django template variables and client-side JavaScript to color-code data maps based on thresholds. The toolbar has an intuitive user interface thanks to its graphical data representation and CSS styling.
Figure 6(a) depicts the hourly air quality index (AQI) for a 24-hour period. The AQI, which ranges from 35 to 90, is represented on the Y-axis, with the X-axis representing the hours of the day from midnight to 11:00 PM. The AQI begins around 45 at midnight, gradually declines until shortly before 06:00, and then rises significantly, peaking at 90 about 09:00. Following this peak, the AQI progressively drops to around 60 by 16:00. In the evening, the AQI rises again, reaching a high of 90 around 22:00 before falling somewhat around 23:00. The day has two main peaks at 09:00 and 22:00, suggesting periods of lower air quality. Air quality tends to improve around midday and early afternoon, maybe due to lower human activity or better meteorological conditions. These data illustrate daily swings in air quality, which might help plan activities to avoid poor air conditions.
Figure 6(b) shows the hourly Air Quality Index (AQI) during peak hours. The AQI values range from 60 to 90 and are plotted on the Y-axis, while the X-axis shows the time of day, which runs from 07:00 to 23:00. The AQI begins at 60 about 07:00 and rapidly rises to a peak of approximately 90 at 09:00. Following this peak, the AQI progressively drops during the morning and early afternoon, reaching approximately 65 by 16:00. The AQI then rises again in the evening, peaking at 90 around 22:00 before significantly falling by 23:00. This trend indicates that air quality is poorest in the early and evening hours, particularly between 09:00 and 22:00, probably due to increased human activity such as driving.
The graphs in Figure 7 display 24-hour moving averages of air quality index (AQI), temperature, and humidity levels recorded hourly during May 2024, covering 744 hours. The AQI fluctuates between 55 and 92, with notable spikes on two specific days: one day when the AQI reached values between 90 and 92, and another with slightly lower AQI values between 83 and 85. These higher AQI readings imply times of worse air quality, but the general pattern shows oscillations between higher and lower values, demonstrating dynamic changes in air quality throughout the month. These fluctuations suggest that air quality may be affected by changing conditions in the laboratory or external environmental factors. Similarly, temperature shows regular variations, with values ranging from 23.0°C to 25.0°C, reflecting diurnal cycles likely related to laboratory activity or external influences. Humidity levels fluctuate between 26% and 30%, with peaks and troughs associated with ventilation systems, air conditioning, or outdoor weather conditions affecting the indoor environment. Overall, the 24-hour moving average smooths out short-term hourly fluctuations and reveals broader patterns in AQI, temperature, and humidity over a month.
On the dashboard page, users can access information about the current temperature, humidity, air quality, and CO2 concentration (see Figure 8).
The EPA has defined the ranges and meanings of each level in the air quality index (AQI). The table in Figure 9 presents the AQI scale and a description of each level. The levels are color-coded for easy identification. Green indicates “good” air quality with an AQI below 50, yellow is “moderate” with an AQI ranging from 51 to 100, orange is considered “unhealthy for sensitive people” with a range of 101 to 150, red indicates “unhealthy” with an AQI of 151-200, purple represents “very unhealthy” with a range of 201 to 300, and anything higher than 301 is classified as "hazardous" and denoted by the color maroon.
As part of the project, an Artificial Intelligence model was implemented to analyze the Apache web server and PostgreSQL database configurations automatically, which allows for the prompt identification of possible vulnerabilities and improved system security. AI generates configuration recommendations by checking access rights, data privacy, and connection security. It should be noted that the recommendations offered by AI are preventive and aimed at improving the overall security of the system and do not necessarily indicate the presence of current problems. Using this approach helps standardize security processes and minimize risks associated with the human factor. In this context, AI improves the efficiency and accuracy of security monitoring, reducing the need for manual verification and accelerating the threat detection process. This approach is especially important in the context of constantly changing cyber threats and the need to maintain the security level of the web infrastructure. In the future, it is planned to expand the model’s functionality by adding support for new algorithms for analyzing complex attacks and integrating with the notification system for prompt notification of detected problems.
This article presents the development of an environmental monitoring system for a laboratory room, which allows for detecting the concentration of air pollutants and monitoring the room's temperature and humidity. Several types of independent sensors were selected, and they were integrated into the Arduino microcontroller. A web server was implemented on the Raspberry Pi platform based on Python to form a single system for measuring, analyzing, and determining the air quality in the room. The module receives parameters such as temperature, relative humidity, ammonia, nitrogen, oxygen, alcohols, aromatic compounds, sulfide, and smoke as AQI and CO2 in real-time at one-hour intervals.
In the future, we will develop an automated system to control ventilation and air conditioning systems automatically. We plan to expand the system by incorporating sensors for PM 2.5, PM 10, and CO and adding a large display for monitoring. Additionally, we will create a user-friendly interface to allow easy interaction with the HVAC system.
The authors have not employed any Generative AI tools. [4] S. M. Dumała, Ł. Guz, and A. Badora, “Indoor Air Quality in Schools Located in Poland, Lublin
Province,” J. Ecol. Eng., vol. 25, no. 1, pp. 17–26, Jan. 2024, doi: 10.12911/22998993/172188. [5] R. S. McLeod, M. Mathew, D. Salman, and C. L. P. Thomas, “An Investigation of Indoor Air Quality in a Recently Refurbished Educational Building,” Front. Built Environ., vol. 7, p. 769761, Jan. 2022, doi: 10.3389/fbuil.2021.769761. [6] E. Zagatti, M. Russo, and M. C. Pietrogrande, “On-Site Monitoring Indoor Air Quality in Schools: A Real-World Investigation to Engage High School Science Students,” J. Chem. Educ., vol. 97, no. 11, pp. 4069–4072, Nov. 2020, doi: 10.1021/acs.jchemed.0c00065. [7] A. Sánchez-Fernández, E. Coll-Aliaga, V. Lerma-Arce, and E. Lorenzo-Sáez, “Evaluation of Different Natural Ventilation Strategies by Monitoring the Indoor Air Quality Using CO2 Sensors,” IJERPH, vol. 20, no. 18, p. 6757, Sep. 2023, doi: 10.3390/ijerph20186757. [8] G. Ulpiani, N. Nazarian, F. Zhang, and C. J. Pettit, “Towards a Living Lab for Enhanced Thermal Comfort and Air Quality: Analyses of Standard Occupancy, Weather Extremes, and COVID-19 Pandemic,” Front. Environ. Sci., vol. 9, p. 725974, Nov. 2021, doi: 10.3389/fenvs.2021.725974. [9] G. Settimo et al., “Indoor Air Quality Levels in Schools: Role of Student Activities and No
Activities,” IJERPH, vol. 17, no. 18, p. 6695, Sep. 2020, doi: 10.3390/ijerph17186695. [10] M. Pulimeno, P. Piscitelli, S. Colazzo, A. Colao, and A. Miani, “Indoor air quality at school and students’ performance: Recommendations of the UNESCO Chair on Health Education and Sustainable Development & the Italian Society of Environmental Medicine (SIMA),” Health Promot Perspect, vol. 10, no. 3, pp. 169–174, Jul. 2020, doi: 10.34172/hpp.2020.29. [11] J. Mazutti et al., “Smart and learning campus as living lab to foster education for sustainable development: an experience with air quality monitoring,” IJSHE, vol. 21, no. 7, pp. 1311–1330, Oct. 2020, doi: 10.1108/IJSHE-01-2020-0016. [12] H. Zhang, R. Srinivasan, and V. Ganesan, “Low Cost, Multi-Pollutant Sensing System Using Raspberry Pi for Indoor Air Quality Monitoring,” Sustainability, vol. 13, no. 1, p. 370, Jan. 2021, doi: 10.3390/su13010370. [13] A. Di Gilio et al., “CO2 concentration monitoring inside educational buildings as a strategic tool to reduce the risk of Sars-CoV-2 airborne transmission,” Environmental Research, vol. 202, p. 111560, Nov. 2021, doi: 10.1016/j.envres.2021.111560. [14] L. Schibuola and C. Tambani, “Indoor environmental quality classification of school environments by monitoring PM and CO2 concentration levels,” Atmospheric Pollution Research, vol. 11, no. 2, pp. 332–342, Feb. 2020, doi: 10.1016/j.apr.2019.11.006. [15] M. Marzouk and M. Atef, “Assessment of Indoor Air Quality in Academic Buildings Using IoT and Deep Learning,” Sustainability, vol. 14, no. 12, p. 7015, Jun. 2022, doi: 10.3390/su14127015. [16] G. Guerrero-Ulloa, A. Andrango-Catota, M. Abad-Alay, M. J. Hornos, and C. RodríguezDomínguez, “Development and Assessment of an Indoor Air Quality Control IoT-Based System,” Electronics, vol. 12, no. 3, p. 608, Jan. 2023, doi: 10.3390/electronics12030608. [17] A. Ramos, V. B. Jesus, C. Gonçalves, F. Caetano, and C. Silveira, “AirVA - Indoor Air Quality Monitoring and Control with Occupants Alerting System,” in Information Systems and Technologies, vol. 801, A. Rocha, H. Adeli, G. Dzemyda, F. Moreira, and V. Colla, Eds., in Lecture Notes in Networks and Systems, vol. 801. , Cham: Springer Nature Switzerland, 2024, pp. 71–81. doi: 10.1007/978-3-031-45648-0_8. [18] X. Dai, W. Shang, J. Liu, M. Xue, and C. Wang, “Achieving better indoor air quality with IoT systems for future buildings: Opportunities and challenges,” Science of The Total Environment, vol. 895, p. 164858, Oct. 2023, doi: 10.1016/j.scitotenv.2023.164858.