In an exclusive interview with Climate Samurai Prof. Raghunathan Rengaswamy, Dean (Global Engagement) & Dept of Chemical Engineering, IIT Madras shared about the mobile air pollution monitoring framework developed by IIT Madras, challanges, technology and more. Here is the excerpt:-
1.Can you provide an overview of the mobile air pollution monitoring framework developed
by IIT Madras? How does it differ from traditional monitoring methods?
Currently, environmental monitoring is done through stations that are fixed at specific locations across the city. Setting up and maintenance of the stations is a significant financial burden on the government. As a result, these stations are distributed sparsely across a city. Studies show that there is a significant spatial variation in air quality within a city. The sparsely distributed, fixed monitoring stations are insufficient to characterise such spatial variation at the necessary spatial resolution. Expanding the number of such stations would be prohibitively expensive. According to a recent news article , TNPCB is planning on setting up 28 stations at a cost of 64 crores. This puts the cost of deployment at roughly INR 2.3 crores per station. The cost of maintaining each station is an additional INR 12-14 lakhs a year. The proposed monitoring solution involves building a network of low cost, mobile environmental monitoring units that are mounted on vehicles that would traverse the entire city. These mobile units would be capable of collecting environmental data with unprecedented spatial resolution. These IoT enabled units would be location aware and capable of bidirectional communication with a central server. This would enable real-time, location-specific tracking of environmental conditions. This high-resolution
data combined with analytical tools would offer deeper insights on location-specific environmental conditions prevailing within a city.
The uniqueness of our device:
- Hyper-local insights are obtained using our low-cost mobile monitoring technology.
- High spatial coverage of road data collection is ensured through our mobile monitoring
technology. - The device features a patented mirror design and structure that can be seamlessly mounted on
vehicles without compromising aesthetics. - State-of-the-art algorithms developed in-house for ensuring the quality of data
- The device combines the functionality of monitoring air quality and road conditions into a single
product, ensuring exceptional precision and delivering a unique, accurate, and comprehensive
solution. - A comprehensive data acquisition pipeline covers the entire process, from collection to storage,
analysis, and visualization.
2. What inspired the development of this low-cost mobile monitoring technology? What challenges were you aiming to address?
Air pollution is one of the major causes for a variety of adverse health effects on human beings. Ambient air pollution is a leading cause of death globally, of which a significant contribution comes from the Indian subcontinent. It has been linked to increased mortality and morbidity. Various epidemiological studies report the ill effects of air pollution on human health . The effects are so conspicuous that it could be considered one of the largest environmental risk factors.
Traditionally, air quality monitoring is performed either to check compliance to national standards or for scientific research. The concentrations of criteria pollutants in the ambient air are periodically measured and reported in the form of Air Quality Index. It involves setting up static monitoring stations with high grade scientific equipment. However, such monitoring stations are generally sparsely distributed across a city. Moreover, they can measure air quality in a small area around their location, depending on their immediate surroundings. Thus, static monitoring stations are able to sample only a small portion of the city’s air quality at a time while leaving out a large part. Furthermore, air quality is extremely dynamic and can vary drastically between locations that are barely a few hundred meters apart . Not only does it vary with space, but it also varies with time in both short and long time-scales. Although fixed monitoring stations are capable of providing highly accurate concentration measurements, they are unable to capture spatial and temporal variability in air quality at high resolutions.
3. How does the IoT-based mobile monitoring framework work? Could you explain the process of gathering spatio-temporal air quality data using sensors mounted on vehicles?
● One potential approach is to utilize existing public vehicles as monitoring agents by equipping them with sensing devices.
● This allows for environmental monitoring with remarkable spatiotemporal granularity, especially when a large number of vehicles are involved.
4. What are the key benefits of using mobile air quality sensors compared to fixed monitoring stations?
● Fixed monitoring stations are limited in their ability to provide specific locality assessments. ● Fixed monitoring stations are unable to capture spatial and temporal variability in air quality at high resolutions.
● For the same number of devices, mobile monitoring can cover a larger area at a higher resolution when compared to fixed devices.
● Mobile devices provide GPS-tagged measurements for air quality information at a range of 1 meter.
● GPS-tagged data can help identify the location at which measurements are being made. This, in turn, allow hotspots to be identified accurately.
| Static Monitoring Mobile Monitoring |
| Results in spatial gaps Spatially continuity in data can be obtained |
| Low coverage efficiency High coverage efficiency |
| High temporal resolution Gaps in temporal information |
| Deployable in any location Restricted to localities with vehicular access |
Kaatru (Air in Tamil) is an ongoing project of the Indian Institute of Technology Madras. The device offers a comprehensive solution by simultaneously monitoring air pollution levels, providing city-wide hyperlocal assessments. One of the key strengths of the device is its hyperlocal accuracy. By capturing real-time data at a granular level, the device enables the precise identification of pollution hotspots. This empowers local authorities to implement targeted interventions and effectively address the root causes of air pollution and road deterioration. The device has an array of sensors that can measure 25 environmental parameters, including ambient air quality, particulate matter, volatile organic compounds, and toxic gases, along with location and time. This multi-parameter profiling approach ensures a comprehensive assessment of air pollution levels and provides a more nuanced understanding of pollution sources. This valuable information can aid in designing effective pollution control strategies and policy frameworks. A standout feature of the device is the personal exposure assessment it offers. By providing individualized information about air pollution levels for each Indian citizen, the device enables people to make informed decisions about their activities and take necessary precautions to minimize their exposure to polluted air. This personalization empowers individuals to protect their health and well-being while contributing to the overall improvement of air quality.
Moreover, the data collected by the monitoring system serves as a powerful tool for evidence-based policy change. By analyzing the insights gained from the device, policymakers and urban planners can identify areas requiring immediate attention and implement targeted interventions. This data-driven approach ensures that policy decisions and mitigation strategies are rooted in accurate and up-to-date information, leading to more effective pollution control measures and sustainable urban development.
The overarching goals of Project Kaatru are:
● Pan-India, hyperlocal, ambient air and road (multiparameter) map
● Personal exposure assessment for each Indian citizen
● Informing policy and mitigation strategies using hyperlocal insights.
The developed technology is the first step in obtaining hyperlocal information at a city level.
6. How does the data collected through this framework enable exposure assessment for Indian citizens? What are the potential implications for personal and public health initiatives?
7. In your research, were there any surprising or noteworthy findings regarding air pollution spikes at specific locations or times? Could you provide some examples?
Association between PM2.5 concetration and anthropogenic activities were established. For instance:
1. A milk distribution hub showed a marked increase between 2 am and 3 am which coincides with the morning milk delivery timings.
2. Distinctive peaks in PM2.5 concentration were observed inschool zones which coincided with school timings.
Similarly, during an short lived extreme pollution event, in this case Diwali, residential areas showed the largest deviation across event and non event days. Similar insights allowed us to associate each land use with its contribution to PM2.5 pollution.
8. How do you envision the practical applications of mobile air quality sensors in various sectors, such as traffic management, government policy, and smart city planning?
- To public- Informed decisions and planning outdoor activities, public health safety ○ Hotspot identification – PM concentrated or intensified regions
- Utilizing air pollution data to inform the strategic placement of hospitals, schools, and residential areas during the planning process.
- Suggesting policy
- Traffic rerouting – Optimizing traffic signal timings based on real-time air quality data to minimize vehicle emissions
9. What parameters can the IoT mobile monitoring devices measure, apart from pollutants? How does the modular design of the devices allow for flexibility and replacement of sensors?
Image below gives the full array of parameters measured by the device
Advantages of modular design
- Plug and sense ie, pre-configured and ready to use upon plugging into a power source
- Easy replacement of sensors. If a new sensor is identified for a different parameter, it could be incorporated into the existing sensor network with minimal downtime.
- Flexibility in sensor selection and configuration
- Scalability is high and cost-effective maintenance
10. Could you explain the patented IoT side view mirror design and its significance in retrofitting the devices onto different types of vehicles?
● Less bulky, compact design, inconspicuous yet functional
● Minimal impact on the aerodynamics of the vehicle
● Non intrusive integration with the side view mirror
11. How is the large volume of data generated by the IoT devices collected, analyzed, and processed using data science principles? What insights can be derived from this analysis?
The data collected by all the devices is aggregated on a central cloud server. Considering the large volume and velocity of the data, big data principles are used to handle and process it. After preliminary processing of the raw sensor data, it is fed to our state-of-the-art algorithms, which are designed in-house specifically to handle spatiotemporal information.
The following insights can be generated from the data:
- Hotspot identification
- Understanding the impact of anthropogenic activities on pollution at specific locations ○ Exposure assessment
- Identifying pollutant concentration along multiple routes
- Impact on nearby cities. ○ Recommendations for policies.
12. Can you provide details about the case studies conducted to validate the framework’s effectiveness? What were the main objectives and findings of these studies?
Two case studies were undertaken as a part of this work. Both the case studies relied on obtaining hyperlocal air quality measurements using a mobile monitoring framework. Both the case studies were run in the south Indian coastal city of Chennai, Tamil Nadu. Each case catered to specific objective, viz.:
1) Regular case: Understanding typical day-to-day spatio-temporal variation in PM2.5 concentration based on anthropogenic activities.
2) Extreme case: Understand an extreme event which is characterized by a pronounced change in PM2.5 concentration over space and in a short time.
Further details can be found in https://www.sciencedirect.com/science/article/pii/S0360132322008277
13. How does the published research paper validate the reliability of the data collected by the IoT devices? Could you discuss the comparison with CPCB stations and the qualitative match observed?
Paper: Data science and IoT based mobile monitoring framework for hyper-local PM2.5 assessment in urban setting
Link : https://www.sciencedirect.com/science/article/pii/S0360132322008277
Laboratory colocation studies were conducted using two of our devices, which had the specifications of sensors against the TSI 3330 Optical Particle Sizer Spectrometer (OPS), over an 18-hour time window and were observed to have a strong correlation between measurements from the devices with good repeatability, indicating reliable data collection. A Pearson’s correlation of higher than 0.97 was obtained, indicating a good match between the two IoT devices and the reference device.
Field collocation with a standardized air quality monitoring station deployed by CPCB demonstrates a qualitative match between IoT device measurements and the reference station, indicating consistency in capturing relative information. Data was collected over 6 days around an extreme event. As seen in the figure below, PM2.5 concentration measurements made by a CPCB station and our mobile device, follow the same profile temporally and across sessions as well. There is, however, a minor quantitative deviation, which can be attributed to various other factors such as the measurement principle, the height at which air is sampled, etc.
Comparison between mobile sensor(top) vs CPCB station(bottom) across 6 days of Diwali
14. What are the future prospects and potential advancements for this mobile air pollution monitoring technology? Are there any plans for scalability or implementation in other regions?
- Traffic and congestion estimation
- Informed city planning
- Personal exposure assessment
- Assessment of health outcomes and interventions
- Source apportionment
- Citywide hyperlocal spatio temporal air quality assessment and hotspot identification
A large scale pilot project is underway in Mumbai, where spatio-temporal profiling of 5 localities will be done over a period of one year. Simultaneously, a high resolution spatio-temporal assessment of entire city of Gurugram has been initiated.
15. Lastly, how do you envision the impact of this technology on air quality management, policy-making, and public awareness regarding air pollution in India?
We believe that this technology offers an affordable means of obtaining highly resolved spatio-temporal air quality, which is not available at present. The availability of such hyperlocal information would enable a plethora of applications for the general public and government bodies. Hyperlocal information is critical for identifying pollution sources and in exposure assessment. TThus,it will enable the design of effective, data driven interventions and mitigation strategies against the detrimental effects of air pollution.
Further, we believe that this technology would augment the existing air quality monitoring infrastructure. Our affordable IoT based mobile monitoring network, coupled with data science principles, offers an unprecedented advantage in gathering hyperlocal insights into air quality. It is the only viable option at present, capable of offering high spatio-temporal awareness that could allow for informed mitigation and policy decisions.