Thank you to our Technical Program Committee for curating this year's program!
Find the full program PDF available here and all conference details available on our conference website. Click each dropdown for oral presentation or poster details.

Oral Presentations

Keynotes

• The 5 Trillion Dollar Question: How can the LCS community achieve maximum impact?
• Presented by: Christa Hasenkopf, Energy Policy Institute at the University of Chicago
  
  Air pollution is one of the largest problems - arguably the largest problem - facing human health, causing an estimated $8.1 trillion USD in health-related damages each year. Yet, the issue is highly under-resourced at the global level, and the resources that do exist are not deployed in many of the countries and regions experiencing the highest health burdens from poor air quality. Why is this the case? And more importantly, what can the air sensing community do about it? The answer to the latter question is: a lot. In this presentation - part analysis, part pep talk, we will delve into how the sensing community can achieve greater impact with its existing resources through adopting and promoting ecosystem-level practices that enable more air quality data to be generated, shared, analyzed, and used for policy without additional resources. It will also explore multiple pathways for how the sensing community is particularly well-positioned to direct the needed attention - and resources - to global air pollution issues, especially where they are most needed. To do this, we will highlight examples that have worked within the field and could be scaled more widely, as well as draw from other successful efforts to address major global health problems, such as HIV/AIDS and malaria.
   
• Next Generation Environmental Sensing
• Presented by: Igor Paprotny, University of Illinois Chicago
   

Innovations in Community Monitoring for Health and Equity

Session chairs: Asim Jaffry, Fair Finance Pakistan and Emil Varghese, CSTEP

• Global Open Air Quality Standards (GO AQS)
• Presented by: Sotirios Papathanasiou, GO AQS
  
  Global Open Air Quality Standards (GO AQS) is a novel framework, addressing the global need for a unified, transparent, and science-backed benchmark for clean air.  Recognizing the limitations of existing air quality indices and the varying regulations across countries, GO AQS aims to provide a universal and easy-to-understand measure of air quality.  The GO IAQS are divided into two tiers:* GO IAQS Starter:  The GO IAQS Starter is designed to be the initial point for improving indoor air quality, specifically aiming to remove financial and logistical obstacles for developing nations and underprivileged communities. Focusing on PM2.5 and CO2, pollutants measurable with widely available low-cost sensor technology.  * GO IAQS Ultimate:  GO IAQS Ultimate is engineered to elevate building performance and occupant health protection by addressing increased levels of pollutants such as O3, CO, CH2O, NO2, and Rn, and meeting more stringent regulatory guidelines.A key feature of GO AQS is the GO IAQS Score, aka the Indoor Air Quality Index (AQI). This simplified index, aligned with color-coded health risk categories enables clear communication of air quality information to the public. The development of GO AQS involved rigorous analysis of breakpoints for each pollutant, drawing extensively on scientific literature and considering their specific health effects at various concentrations.The GO AQS initiative is steered by a diverse committee composed of scientists, public health professionals, and air quality experts. This democratic structure, with equal voting power among members, ensures broad consensus and accountability in establishing and maintaining these vital air quality standards.  
   
• Pollution from Coal Power Driving Public Health Inequities and Death
• Presented by: Asim Jaffry, Fair Finance Pakistan
  
  Rising cases of strokes, asthma, and other health issues are linked to elevated PM2.5, nitrogen, and ozone levels near coal-fired power plants in Pakistan. Fossil fuels—coal, oil, and natural gas—constitute 60 percent of the country's electricity capacity. This talk examines the health impacts of coal pollution and provides evidence to guide financial institutions and international financial institutions (IFIs) in reducing air pollution. It highlights the role of the private sector in advancing the transition to cleaner energy. Approximately 44 million people—about one-fifth of the population—live within 50 km of fossil fuel power plants in Pakistan, facing increased health risks. In these areas, children suffer from chronic asthma, the elderly are at higher risk of strokes and heart failure, and pregnant women face complications due to toxic air pollutants. Emissions of sulfur dioxide (SO₂), nitrogen oxides (NOx), and fine particulate matter harm marginalized communities grappling with limited healthcare access and economic disparity. Over the past two decades, there has been a surge in non-communicable diseases (NCDs)—such as heart disease and chronic lung conditions—where air pollution is a major risk factor. These NCDs account for over half of all deaths in the country. Using the Jamshoro coal-fired plant as a case study, the session will demonstrate that supercritical technology, while slightly more efficient, does not effectively reduce hazardous emissions and public health risks. The session will assess Pakistan’s regulatory framework regarding air pollution, calling for better emission reduction strategies and interventions to combat fossil fuel emissions and cross-border air pollution. The lack of greenhouse gas (GHG) standards and minimal private sector engagement complicates the transition away from fossil fuels. Air pollution is a solvable crisis. Recommendations will include enhancing regional cooperation, imposing strict emissions criteria, and increasing climate finance and technical support to move beyond fossil fuels. These measures are essential for protecting public health and ensuring access to clean air for all.
   
• Clearing the Air: Pathways to a Just Energy Transition in Asia Pacific
• Presented by: Shrishty Anand, Fair Finance Asia
  
  Transitioning from fossil fuels to renewable energy is crucial for addressing climate change and air pollution. However, this shift must be equitable, ensuring that vulnerable communities—particularly coal-dependent workers, women, and marginalized groups—are not further disadvantaged. This presentation examines the financial, social, and policy aspects of a Just Energy Transition (JET) in the Asia-Pacific region. The region remains heavily reliant on coal, with China, India, Indonesia, and Vietnam accounting for a significant portion of global coal consumption. Continued financing of coal projects by regional banks, despite their growing commitments to environmental, social, and governance (ESG) criteria, undermines climate objectives and hampers progress. A truly just transition requires not only divestment fromfossil fuels but also strategic investments in renewable energy, alongside robust support systems for affected communities. This includes retraining programs, alternative income sources, and strong social safety nets. Additionally, the energy transition is inherently gendered. Women in the region disproportionately experience energy poverty, are underrepresented in clean energy jobs, and bear increased care responsibilities due to ecological stress. Without intentional policies that prioritize gender inclusion—such as gender-responsive financing, leadership opportunities, and equitable access to green employment—the transition could reinforce existing systemic inequalities. A just energy transition in the Asia-Pacific is both a climate necessity and a moral imperative. Aligning capital flows with inclusive policy frameworks and embedding gender equity at every level will empower the region to build a resilient economy. It will also foster a people-centered, green, and gender-transformative JET in Asia. Insights from Fair Finance Asia provide actionable strategies for policymakers, financiers, and advocates to ensure an equitable shift to clean energy—making sure that no one is left behind. The Asia-Pacific region, home to some of the world’s fastest-growing economies, faces an urgent challenge: reconciling rapid industrialization with the critical need to decarbonize. The Asian Development Bank estimates that 4 billion people, or 92% of the population in the Asia-Pacific region, are exposed to air pollution levels that pose significant health risks. Air pollution and greenhouse gas (GHG) emissions present intertwined challenges that threaten our climate, health, and economy.
   

Policy, Funding, Expanding Data Networks and Sustainability for Air Quality Monitoring

Session chairs: Anil Namdeo, Northumbria University and Elizabeth Vega, UNAM Mexico

• Role of Philanthropy in advancing Low-cost AQ sensors for policy use
• Presented by: Shriram Manogaran, Clean Air Fund
  
  The Clean Air Fund (CAF) has successfully advanced its Low-Cost Sensor (LCS) strategy, achieving its short-term goals of improving accuracy and reducing costs through grantee initiatives. The deployment of LCS across various settings has resulted in increased availability of local air quality data, both through direct grantee efforts and external scaling. These advancements have been realized relatively quickly since CAF’s overarching data strategy launched in 2022.CAF’s impact extends to policy and practice, with city governments—such as those engaged through Breathe London, C40 Cities, and Breathe Accra—incorporating LCS data into air quality (AQ) policies. This validates CAF’s assumption that enhanced accuracy and affordability of LCS drive adoption, albeit in ways that are sometimes intangible. The evolving LCS landscape has also highlighted key challenges, including the need for improved source identification, measurement of pollutants beyond PM2.5, and addressing operational costs, which extend beyond the initial purchase of sensors. Additionally, while city municipalities have integrated LCS with reference-grade monitors, there has been limited adoption by community-based organizations (CBOs), NGOs, and smaller cities.CAF’s strategic added value lies in its unique ability to integrate funding with in-house technical expertise, act as a networking and convening force across diverse stakeholders, and support an innovation-to-scale pipeline. Rather than funding long-term implementation, CAF prioritizes testing, piloting, and demonstrating innovative AQ solutions. Its approach is guided by a focus on areas with the greatest need and opportunity, emphasizing local knowledge and capacity-building to ensure sustainability. Moreover, CAF plays a critical role in early-stage field building by strategically supporting LCS development through innovation, gap-filling, and fostering collaborative networks. This comprehensive approach positions CAF as a key driver in advancing AQ monitoring and policy integration globally.
   
• Developing Bangkok’s First Clean Air Management Plan: A Roadmap for Sustainable Air Quality Improvement
• Presented by: Pongsatorn Greigarn, C40 Cities Climate Leadership, Inc
  
  Effective air quality management requires a comprehensive monitoring framework and a clear policy roadmap to drive sustainable improvements. In Bangkok, the Breathe Cities initiative is supporting the development of its first-ever long-term Clean Air Management Plan to reduce PM2.5 levels and align with Thailand’s national air quality standards and World Health Organization Air Quality Guidelines. This multi-year plan will integrate air quality monitoring, emissions controls, stakeholder engagement to transition Bangkok towards sustainable clean air management. It will align with the forthcoming national Clean Air Management Act, incorporate regional best practices, and strengthen institutional capacity for effective policy implementation. The Clean Air Management Plan will also mainstream sectoral policies, prioritize public health outcomes, and promote inclusive governance, ensuring participation from vulnerable communities and key stakeholders. To sustain long-term air quality improvements, the plan will enhance monitoring, evaluation, and enforcement mechanisms while leveraging Bangkok’s strong policy commitment and dedicated funding for air quality infrastructure. By expanding real-time monitoring networks, including low-cost sensors, Bangkok aims to bridge data gaps, inform targeted interventions, and improve  public health outcomes. This presentation will explore key components of Bangkok’s approach to develop the Clean Air Management Plan, highlighting the intersection of air quality monitoring and policy development. It will also emphasize the importance of data-driven decision-making, inclusive governance and strategic interventions to create a scalable model for effective air quality management in urban environments.
   
• From Data to Action: How Air Quality Monitoring Secured $50M in Funding for Pollution Reduction in Central Asia
• Presented by: Ethel Garcia, Clarity Movement
  
  Accurate air quality data is a critical tool for driving policy change and securing investment in pollution mitigation efforts. The Almaty-Bishkek Economic Corridor (ABEC) Air Quality Initiative exemplifies how a focused approach to air quality monitoring can catalyze large-scale funding and action. By deploying over 100 Clarity Node-S air quality sensors across Almaty and Bishkek, local governments established a robust, real-time dataset on particulate matter pollution, providing the foundation for evidence-based policy recommendations and investment strategies.The data collected from these sensors enabled a series of high-impact analyses, including a World Bank study assessing PM₂.₅ pollution sources in Bishkek. This report identified key contributors to air pollution and modeled the potential impact of various mitigation measures, helping to build a strong case for targeted policy interventions. By demonstrating the connection between air quality improvements and broader benefits such as reduced greenhouse gas emissions and improved public health, the analysis played a pivotal role in securing over $50 million in funding from the World Bank for air quality improvement projects. These investments will support measures such as clean heating solutions, urban greening, and strengthened air quality management systems.This session will showcase how data-driven air quality monitoring can translate into policy impact and financial investment, using the ABEC Air Quality Initiative as a case study. We will explore how Clarity’s sensor network and advanced data analysis addressed critical information gaps, how the findings were leveraged to influence decision-makers, and how similar approaches can be applied in Southeast Asia to unlock funding for sustainable air quality improvements.
   

Data Analytics, Trends, and Forecasting Technologies

Session chairs: Shriram Manogaran, Clean Air Fund and Vasu Kilaru, US EPA

• AI-enriched automation for evaluating health risks from air pollution
• Presented by: Andrei Dusmikeev, BreatheSafetyIndex
  
  Traditional air quality indices, such as the EPA’s Air Quality Index (AQI), provide generalized insights but fail to account for individual health vulnerabilities, activity levels, and exposure patterns. The increasing urbanization and adoption of digital health tools demand automated, real-time health risk assessments for personalized and community-level decision-making. This study presents BreatheSafetyIndex, an AI-powered, API-driven automation framework [1] that transforms underprepared air quality data into precise, actionable health risk evaluations.
   
• Multilinear Regression Analysis of PM2.5 Concentration and Meteorological Factors in the Kampala and Fort Portal City
• Presented by: Fidel Raja Wabinyai, AirQo
  
  This study examines the relationship between meteorological factors, lagged PM2.5 concentrations, and PM2.5 dynamics in Kampala and Fort Portal, Uganda. Using multilinear regression models (Ordinary Least Squares (OLS)), we analyze seasonal variations and highlight the critical role of meteorological parameters—temperature, humidity, wind speed, and direction—alongside prior day PM2.5 levels. Results reveal substantial differences between the two cities, with Kampala, characterized by urbanization and industrial activity, reporting higher PM2.5 levels (dry season mean: 38.68 µg/m³) compared to Fort Portal (dry season mean: 27.53 µg/m³). Model performance varied across seasons, achieving higher explanatory power during dry periods (R² = 0.57 in Kampala; R² = 0.71 in Fort Portal). Wet seasons demonstrated reduced predictability due to additional variables like rainfall and atmospheric dynamics.The study highlights the persistence of pollution during dry seasons in both cities, attributed to limited dispersion and pollutant accumulation. Statistical diagnostics, including Durbin-Watson statistics (1.7-2.2), confirm the robustness of our models, with minimal autocorrelation. Fort Portal's smaller, agriculture-based economy exhibited less pollution variability than Kampala, where urban dynamics intensify air quality challenges.The findings underscore the significance of meteorological variability and lagged PM2.5 in shaping pollution patterns across the two cities, offering actionable insights for public health interventions and pollution control strategies. The study advocates integrating additional variables, such as rainfall intensity and localized pollution sources, to enhance predictive accuracy and inform evidence-based policymaking. These insights are critical for advancing air quality management in rapidly urbanizing regions.
   
• Use of Intelligent Low Powered Devices for Air Quality Monitoring & Understanding Sensor Health
• Presented by: Prithviraj Pramanik, National Institute for Technology Durgapur
  
  Poor air quality is a menace in developing countries, with the situation being particularly dire in Southeast Asia. Thus, air quality monitoring is crucial to understanding the impact on public health, especially in regions where air quality is hazardous. Traditional air quality sensors are often very expensive to purchase and require continuous maintenance, which adds to the cost burden. This poses a serious challenge for widespread deployment. The current alternative is low-cost sensors, which offer a solution but suffer from reliability issues such as sensor drift and sensitivity drift. Thus, to correct the issues, in a large-scale deployment, the cost burden can often rise and lead to a condition where the cost of deploying low-cost sensors is not low at all. To address these issues, we propose using TinyML based on a sensor data processing system to enhance the reliability and accuracy of low-cost air quality sensors.Implementing the TinyML stack directly on the sensors enables the deployment of machine-learning models on microcontrollers and edge devices, thus allowing real-time data processing and analysis. In addition, we utilize meteorological factors in real life, which is especially relevant in South Asia and Southeast Asia, where, in the tropics, most low-cost sensors, especially PM2.5, are heavily influenced by humidity. Using TinyMl, we are implementing machine learning algorithms to create models that correct for the sensor drift, filter out noise and detect when the sensor is behaving very erratically, which helps us improve the overall performance of the low-cost sensors while giving us the scope to understand "sensor health." These models are trained on a combination of historical sensor data and real-time inputs from sensors measuring PM2.5, meteorological factors, etc. The model ensures that the sensor provides more accurate and reliable data, critical for effective air quality monitoring.Additionally, processing the data at the edge using TinyML reduces the dependency on cloud-based resources, thus further lowering long-term operational costs without improved latency. This makes the solution particularly suitable for deployment in remote and underserved regions with limited internet access. We have initiated this work for both outdoor and indoor settings and plan to validate the approach in various settings and compare the performance of our model with that of traditional high-cost sensors. Initial results show a high correlation, thus warranting a broader study.We aim to address the critical gap of a robust framework for validating low-cost "sensor health" through this study. Not only is it essential for regulatory evidence, as seen in some of the major countries in the subcontinent, but it is also essential to understand the actual health burden.
   
• Transforming Air Quality Data with GIS and AI: Integrated Analysis for Scalable Environmental Solutions
• Presented by: Kruti Davda, Oizom Instruments Private Limited
  
  The increasing complexity and volume of air quality data present significant challenges for effective environmental management. This presentation introduces an advanced GIS-based platform developed by Oizom, integrating artificial intelligence (AI) and machine learning (ML) algorithms to transform raw sensor and environmental datasets into comprehensive analytical outputs. By combining air quality data with weather parameters, vegetation indices, hydrological features, and satellite-derived fire hotspots, the platform enables multi-dimensional trend analysis, source attribution, and predictive insights.The system employs AI-driven methodologies to automate data interpretation, detect anomalies, and uncover hidden patterns in pollution trends. These capabilities enhance the accuracy and scalability of real-time monitoring while supporting targeted interventions and long-term strategic planning. A case study from North India demonstrates the tool’s ability to generate high-resolution daily pollution gradient maps and analyze correlations between environmental factors and air quality dynamics.This session will explore the innovative architecture of the platform, highlighting its use of GIS and AI technologies to bridge the gap between raw data and actionable insights. Key advancements include scalable real-time analytics, integration of multi-source datasets, and applications in urban planning, environmental forecasting, and sustainable development.
   

Public Health and Environmental Impacts with Air Sensors

Session chairs: Shih-Chun Candice Lung, Academia Sinica, Taiwan, and To Thi Hien, University of Science, Vietnam

• Quantifying Long-Term Indoor Fine Particulate Matter (PM2.5) Exposures Using Real Time Low-Cost Sensors in an Urban Cohort of Pregnant Women in Southern India
• Presented by: Naveen Puttaswamy, Sri Ramachandra Institute of Higher Education and Research
  
  Exposure to fine particulate matter (PM2.5) during pregnancy can negatively impact fetal development and birth outcomes. Characterizing long-term exposures throughout pregnancy is crucial for identifying critical windows of exposure related to adverse birth outcomes. Low-cost sensors (LCS) offer continuous PM monitoring solutions with high temporal resolution, essential for exposure - response analysis. Our aim is to characterize long-term PM2.5 exposures using LCS in an urban cohort of pregnant women in southern India.Methods: As pregnant women spend most of their time indoors, we used real-time PM sensors equipped with PMS7003™ (Plantower Inc., China) to monitor living-room PM2.5 levels (as proxy for their personal exposures) for 6 to 8 months in a subset of 15 homes within the ongoing REACH (Reproductive effect from Exposures to Airborne Chemicals in urban Homes) cohort (n=300) in Chennai, India. The LCS records PM, temperature, and relative humidity at 1-minute time intervals, transmitting data in real-time to the cloud servers. Sensors were collocated with reference-grade monitor (BAM1022) before and after each home measurement, and linear models were trained using collocated data to derive ambient calibration coefficients.Results: To date, continuous PM2.5 data has been monitored for an average of 160 ± 30 days across 15 homes in the REACH cohort, with data availability ranging from 82% to 99%. Sensor precision was satisfactory, with a standard deviation between 6.0 and 9.8 µg/m3 and a coefficient of variation from 8.8 to 21.2%. The normalized root mean square error (NRMSE) for ambient collocation varied from 29.5 to 48.5% (r = 0.3 to 0.7). Long-term averages of indoor PM2.5 levels were higher on weekends (19.2 to 89.9 µg/m3) compared to weekdays (13.4 to 48.6 µg/m3). Diurnal plots indicated that indoor PM2.5 levels peaked twice daily: from 7 to 10 a.m. and 6 to 10 p.m. Elevated indoor PM2.5 levels were particularly noticeable during the evening window across all homes.Summary: We demonstrate the applicability of LCS for long-term indoor PM monitoring in cohort studies. We will present space-time variability in indoor PM2.5 levels over the pregnancy period and its influence on health risks associated with indoor air pollution.
   
• Application of low-cost PM sensors in the Philippines: Insights from two contextually-unique personal monitoring studies
• Presented by: Maria Obiminda Cambaliza, Ateneo de Manila University
  
  Metro Manila, Philippines is a megacity with a population density of roughly 23,000/km2.  As an emerging economy, the transportation sector contributes more than 80% to pollution emissions in the city.  While fixed-site measurements of PM2.5 at an urban-mixed station in Metro Manila showed a decreasing trend over recent years, there are sectors in the society that remain at high health-risk due to their continued exposure to elevated levels of particulate pollution.  This presentation focuses on the application of low-cost PM2.5 sensors in the Philippine context, specifically in Metro Manila.  We present results of our measurement and analyses for two cohorts: (1) public utility jeepney (PUJ) drivers, a high-risk occupational group in the country, and (2) a group of university students and employees.  Our comprehensive measurements showed that the personal exposure of PUJ drivers, at 36 µg/m3, was twice as much as the mean exposure of the general population (~18 µg/m3 determined from the urban mixed site).  This contrasts greatly with the mean exposure of university employees and students, which was revealed to be lower, i.e. about 2/3 of the mean annual PM2.5 concentration at an urban mixed site.  However, personal monitoring of this group further revealed the significant contribution of indoor microenvironments (homes, restaurants, cafeteria, etc.) to their exposure to particulate pollution.  These two contextually-unique studies highlight the advantages of portable, low-cost sensors in quantifying the personal exposure of individuals with temporally and spatially-diverse activity patterns.
   
• Assessing PM2.5 Sources, Exposure, and Health Impacts in Mandalay's Exposed Communities: The Role of Low-Cost Air Sensors
• Presented by: Ohnmar MayTin Hlaing, Environmental Quality Management Co., Ltd
  
  Nearly 99% of the global population is exposed to air pollution levels exceeding WHO's PM2.5 guideline, with the highest exposure in low- and middle-income countries. This widespread air pollution contributes to approximately 7 million premature deaths annually, primarily from cardiovascular and respiratory conditions (WHO, 2025). In Myanmar, air pollution accounts for an age-standardized mortality rate of 176 deaths per 100,000 (2019), highlighting a critical public health issue.Therefore, conducting research that provides policy-relevant insights is essential for mitigating PM2.5-related health risks at the national level. However, in developing countries like Myanmar, where air monitoring infrastructure is still emerging and high-end monitoring equipment is limited, low-cost air sensors become essential. They offer an effective solution for evaluating PM2.5 sources, exposures, and the exposure-health relationship, providing high temporal and spatial resolution at significantly lower costs. This study aimed to assess environmental (ambient), personal PM2.5 exposure and associated health impacts in communities in Mandalay using low-cost sensing (LCS) devices. PM2.5 concentrations were monitored over one year, covering summer, rainy, and winter seasons, using AS-LUNG-O and AS-LUNG-P monitors, while air pollution dispersion was modeled via AERMOD. Key findings indicated significant seasonal and spatial variations in PM2.5 concentrations. In the town area, the highest levels were recorded in summer (90 ± 18 µg/m³) and winter (36 ± 7 µg/m³), primarily due to mobile emissions, while the lowest levels were observed during the rainy season (16 ± 2 µg/m³) due to the washout effect. In the periurban waste disposal area, summer levels reached 194 ± 19 µg/m³ due to open waste burning, with lower levels in the rainy (61 ± 11 µg/m³) and winter (68 ± 11 µg/m³) seasons, attributed to dilution effects and the cessation of open burning respectively. In terms of spatial variations, the most highly polluted area was the periurban waste disposal site which is about 3 km to the northeast from the town. The simulated pollutant levels in the town were not remarkably impacted by the disposal site due to upwind location. Personal exposure measurements revealed that waste handling workers experienced significantly higher PM2.5 exposure (140 ± 4 µg/m³) compared to office-based controls (90 ± 2 µg/m³), exceeding Myanmar’s National Environmental Quality Emission Guideline (25 µg/m³). The highest recorded exposure was 152 ± 52 µg/m³ during 30-minute intervals at the waste disposal site. Health indicators, including Blood Pressure (BP), Heart Rate (HR), Peripheral Oxygen Saturation (SpO₂), and Peak Expiratory Flow Rate (PEFR), were monitored to identify potential health impacts linked to PM2.5 exposure. It was found that heart rate, one of the key health indicators, is more closely associated with PM2.5 exposure compared to other factors. Future research on the correlation between PM2.5 exposure and HR/HRV is underway to identify the health damage coefficients which determine how exposure to fine particulate matter affects human health, which can be used for policymaking and public health interventions. The study highlights the role of LCS devices in identifying pollution hotspots, understanding exposure disparities, and guiding policy recommendations for effective air pollution control.
   
• Source evaluation, personal exposure assessment, and epidemiological analysis in Asia using research-grade low-cost PM sensors
• Presented by: Shih-Chun Candice Lung, Academia Sinica, Taiwan
  
  The development of low-cost sensors has opened new opportunities in aerosol research. By calibrating these sensors in both laboratory and field settings using research-grade instruments, concerns regarding data accuracy are effectively addressed. These research-grade low-cost sensors enable the measurement of PM2.5 and PM1 with minute-level resolution. This presentation highlights their application in evaluating community and indoor PM pollution sources, assessing personal PM exposure, and conducting panel-type epidemiological studies that examine the relationship between peak PM exposure and heart rate variability (HRV). As an indicator of cardiac autonomic balance, reduced HRV has been linked to an increased risk of myocardial infarction. Case studies from Taipei, Nantou, and Kaohsiung in Taiwan and Ulaanbaatar, Mongolia will be presented as examples. The application of sensors to evaluate community PM sources was conducted in Nantou, Taiwan in 2017, while their use in assessing indoor PM sources took place in the Taipei metropolitan area in northern Taiwan in 2018. PM exposure assessments and panel-type epidemiological studies were carried out in Kaohsiung, Taiwan in 2018 and 2020. A study in Ulaanbaatar, Mongolia focusing on PM exposure assessment and exposure-health evaluation is on-going since 2024. Research-grade low-cost sensors—AS-LUNG-O, AS-LUNG-I, and AS-LUNG-P—were utilized for outdoor, indoor, and personal monitoring, respectively. Additionally, medically certified RootiRx® sensors were employed for HRV monitoring. The results showed that incremental contributions from the stop-and-go traffic, market, temple, and fried chicken vendor to PM2.5 levels at 3-5m away were 4.38, 3.90, 2.72, and 1.80 μg/m3, respectively. Significant PM spatial variations observed further emphasized the importance of conducting community air quality assessment. For indoor sources, cooking occurred most frequently; cooking with and without solid fuel contributed to high PM2.5 increments of 76.5 and 183.8 μg/m3 (1 min), respectively. Incense burning had the highest mean PM2.5 indoor/outdoor (1.44 ± 1.44) ratios at home and on average the highest 5-min PM2.5 increments (15.0 μg/m3) to indoor levels, among all single sources. In exposure assessment and epidemiological studies, it was found that for a 10 μg/m3 increase in PM2.5, HRV indicators were reduced 1.3-4.0% in Taiwan subjects in summer. In Ulaanbaatar, Mongolia, 20 subjects were recruited with significant variations in PM exposure. The low-cost sensors used and methodology demonstrated in this presentation can be applied in resource-limited countries to conduct PM and health research.
   

Adapting Low-Cost Sensors for Southeast Asian Environments

Session chairs: Everlyn Tamayo and Carlo Bontia, Clean Air Asia

• Introduction to Humidity-Resilient, Low-Cost Sensor Solution for PM monitoring
• Presented by: David Riallant, Groupe Tera
  
  The rapid urbanization and industrialization of Southeast Asia have raised growing concerns about air quality, particularly regarding fine particulate matter (PM1, PM2.5, and PM10).While low-cost sensor (LCS) technologies offer a scalable solution for widespread air quality monitoring, their deployment in this region presents unique challenges—primarily due to high humidity levels that compromise data reliability, leading to inaccuracies, sensor drift, and premature aging.In this talk, we will introduce a next-generation optical particulate matter sensor specifically designed to overcome these environmental constraints. Unlike conventional low-cost sensors, which suffer from hygroscopic particle growth and degradation over time, the NextPM sensor integrates a patented humidity regulation system. This innovation ensures stable and accurate measurements even in high relative humidity conditions, preventing data drift and sensor clogging. Combined with its dual detection angle (measuring light scattering at two angles, 45° and 90° ) NextPM also improves measurement precision to offer an overall accurate, reliable and yet affordable dust monitoring solution.We will present a comparative analysis of sensor performance—with and without humidity regulation—against reference instruments, along with statistical insights demonstrating the effectiveness of this technology in humid environments. NextPM has been granted Class 1 Certification in Korea, a region with similarly high humidity conditions. With an accuracy of 80% or higher compared to official reference instruments, the sensor is validated for use in both regulatory and scientific applications, offering a reliable solution for improved air quality monitoring in challenging climates.
   
• Nanophotonic-Based Air Quality Sensors for Tropical Environments: Exploring Novel Materials for Enhanced Performance
• Presented by: Sandeep Battula, Qualivon Technologies Pvt Ltd
  
  Southeast Asia's tropical climate poses significant challenges to air quality monitoring, including high humidity, extreme temperatures, and diverse pollution sources. Traditional sensing technologies often struggle to maintain accuracy and reliability in these conditions. This presentation introduces a novel approach to air quality monitoring using nanophotonic-based sensing technologies. This research discusses about the selection of materials, design, fabrication, and characterization of nanophotonic structures that leverage the unique properties of novel materials (Silicon Nitride, Lithium Niobate, Indium Phosphide, & Germanium on Silicon) to enhance sensor performance. These materials have been specifically selected for their ability to withstand the harsh conditions of tropical environments, while maintaining high sensitivity and selectivity to target pollutants. Some of the sensing photonic structures that we want to discuss includes the photonic crystals, microring resonators, on-chip Mach Zender Interferometers & Sub wavelength Grating structures.Key results will highlight the improved accuracy, reliability, and durability of our nanophotonic-based sensors at lower costs in tropical environments. In this research we will be also showcasing our tested nanophotonic structures for gaseous like NH3, SO2, NO and our current work on CH4 sensing. We will also discuss the potential applications of this technology in urban and rural Southeast Asia, including integration with existing monitoring networks and deployment in areas with limited infrastructure. By exploring new materials and technologies, this research aims to contribute to the development of robust, cost-effective, and sustainable air quality monitoring solutions for Southeast Asia.
   
• Tracking Air Pollution on the Move using Mobile Micro-Sensors
• Presented by: Adisorn Lertsinsrubtavee, Asian Institute of Technology
  
  Air pollution, particularly PM₂.₅, poses significant health risks to outdoor workers in urban areas such as Bangkok and Chiang Mai. However, their actual exposure remains poorly understood, as traditional fixed-site monitors often underestimate real-world conditions. To address this gap, we introduce MobileSense, a low-cost, mobile air quality monitoring platform that provides real-time PM₂.₅ data through a helmet-mounted sensor system equipped with GPS tracking and alerts. A seven-month field study revealed that frontline motorcycle riders experience significantly higher PM₂.₅ exposure than recorded by fixed monitors, often exceeding WHO safety limits—even during non-polluted seasons. The spatiotemporal analysis of PM₂.₅ distribution across both cities identifies high-risk areas and peak pollution periods, offering critical insights for air quality management and worker protection.
   
• Black Carbon Measurement: A Powerful Tool for Better Particulate Matter Management in Southeast Asia
• Presented by: Jack Kodros, Clarity Movement
  
  Black Carbon (BC) is a key indicator of combustion-related air pollution, with significant health and climate implications. Despite its importance, widespread, affordable, and reliable BC monitoring solutions remain scarce, particularly in Southeast Asia, where pollution sources vary widely due to urbanization, biomass burning, and transportation emissions. This presentation explores the integration of Clarity’s novel BC Module into a low-cost sensor (LCS) platform and its potential applications for air quality management in the region.Long-term collocated measurements in diverse environmental conditions across California, Colorado, and Florida demonstrate that the Clarity BC Module achieves high accuracy (R² > 0.8) compared to research-grade aethalometers. Expanding on this work, recent deployments in Perth, Australia, supplementing an existing Clarity PM₂.₅ network, have shown that BC measurements provide a more detailed picture of combustion sources, highlighting localized pollution hotspots. By applying a source apportionment model, we found that BC variability—especially from fossil fuel combustion—was more pronounced than that of PM₂.₅, making it a valuable tool for identifying traffic-related emissions. Additionally, during periods impacted by regional biomass burning, BC levels closely correlated with PM₂.₅, reinforcing its role in distinguishing combustion sources.Southeast Asia presents unique challenges for air quality monitoring, including high humidity, seasonal haze events, and rapidly changing pollution patterns. Integrating BC monitoring with existing PM₂.₅ sensor networks can provide deeper insights into pollution sources, enabling more effective interventions to mitigate emissions. With increasing regulatory focus on air pollution control in the region, adopting BC measurements alongside LCS networks can empower policymakers, researchers, and communities to make data-driven decisions to improve air quality.This session will explore how low-cost BC sensors can enhance air pollution monitoring and management strategies across urban and rural Southeast Asia. We will discuss practical deployment considerations, lessons learned from existing networks, and opportunities for integrating BC measurements into regional air quality initiatives. By leveraging the affordability and scalability of LCS, cities, and communities can improve their understanding of pollution sources, implement targeted mitigation strategies, and work toward cleaner air for all.
   
• Updated Exposure Estimate for Indonesian Peatland Fire Smoke using Network of Low-cost PurpleAir PM2.5 sensors
• Presented by: James McQuaid, University of Leeds
  
  Air pollutant emissions from wildfires on Indonesian peatlands lead to poor regional air quality across south-east Asia. Fine particulate matter (PM2.5) emissions are particularly high for peat fires leading to substantial population exposure to PM2.5. Despite this, air quality monitoring is limited in regions close to peat fires meaning the impacts of peatland fires on air quality is poorly understood and it is difficult to evaluate predictions from atmospheric chemistry models. To address this, we deployed a network of low-cost (Purple Air) PM2.5 sensors at 8 locations across Central Kalimantan, where peat fires are frequent. The sensors measured indoor and outdoor PM2.5 concentrations during August to December 2023. During the haze season (September 1st to October 31st), daily mean outdoor concentrations were 120 mg m-3 but peaked at >400 mg m-3. Indoor PM2.5 concentrations were only ~10% lower (mean 110 mg m-3), indicating that is difficult for the population to reduce their exposure to PM2.5 from fires. The reduction in mean PM2.5 concentrations between outdoor and indoor environments was larger in urban locations (-11%) compared with rural locations (-3%), suggesting urban housing may provide better protection from outdoor air pollution. To generate an updated assessment for the population’s exposure to peatland fire PM2.5 we combine the information from monitoring both indoor and outdoor PM2.5 concentrations with modelled ambient (outdoor) PM2.5 concentrations from the WRF-Chem atmospheric chemistry transport model. Our updated exposure assessment accounts for the population’s personal exposure to peatland fire PM2.5 for the first time.
   

Infrastructural Challenges and Needs in the Air Sensors Domain

Session chairs: Shriram Manogaran, Clean Air Fund, Vasu Kilaru, US EPA and Sebastian Diez, Universidad del Desarrollo

• Using 101 Colocations from 39 Cities to Create a Global Correction for Optical Low Cost PM2.5 Sensors
• Presented by: Garima Raheja, Columbia University
  
  Air pollution is a leading cause of global premature mortality. Traditional methods of measuring air pollution are expensive, technically challenging, and inaccessible for many low-income marginalized communities. Advancements in low-cost sensors (LCS) are helping bridge the data gap left by these reference-grade monitors. Novel data science techniques are being used to develop correction factors for LCS, but these studies generally 1. use co-locations with expensive reference-grade monitors 2. utilize temperature, humidity and other measurements to account for variation in hygroscopicity and optical properties and 3. are often local in scope, limited to one city or metro area.Can we use correction factors developed in one community, in another? We use colocations contributed by 39 community projects and regulatory studies (including in NYC, Ohio, Accra, Lomé, Kinshasa, London, and Kolkata) at varying climatologies to assess the performance of 4 machine learning techniques, and compare them to correction factors in the literature. Additionally, we develop a Global Gaussian Mixture Regression (GMR) machine learning model trained on co-locations from communities in the Clean Air Monitoring and Solutions Network (CAMS-Net). We find that in most cases, the Global Gaussian Mixture Regression model performs 57-96% as well as using a local correction model, which means that using this model could provide high levels of accuracy in a community using LCS without the need for a $100,000+ reference monitor colocation. In some cases, such as Nairobi, the Global Gaussian Mixture Regression model is actually better than using correction developed with local colocation.Global GMR is greater than the sum of its parts: contribution from some communities has reciprocated progress in many more. We present an open-source dashboard that enables the correction of data from 20,000+ PurpleAir and Clarity sensors around the world without a reference monitor colocation, and has allowed community groups, regulators and policymakers around the world to make the most of their LCS data.
   
• Effective management of air quality data: The need, a case study, and a current project to build customizable open-source data management system (DMS)
• Presented by: Carlo Bontia, Clean Air Asia
  
  Air quality data is crucial in informing air quality management decisions, yet majority of countries have yet to meet the recommended minimum number of monitoring stations relative to their population. Air quality sensor networks provide an opportunity to fill data gaps at a lower cost than reference stations. As more governments and organizations employ and plan for high-density hybrid monitoring networks, there is a need to address challenges and explore solutions related to the (a) procurement, maintenance, and operation of sensor networks; (b) effectively communicating real-time air pollution data; (c) explaining and using air pollution data for policies and actions; and (d) developing and implementing an efficient air quality data management system (DMS), which is an essential part of any monitoring network to efficiently manage data, aiding in the analysis and implementation of measures to reduce air pollution.We present opportunities to address sensor monitoring network challenges, building on Clean Air Asia’s extensive network of national and local governments in the region and hands-on work in Metro Manila, Philippines. Solutions include important aspects of intensive and sustained capacity building, efficient procurement processes, and operational planning for monitoring and air quality communication and policy development. Further, we highlight key considerations in developing an air quality DMS that match the needs, resources, technology, and technical capacity of stakeholders. We present a case study on the use of a customized Python-based air quality data analysis tool built on Google Colabs, demonstrating the increase in analysis efficiency. We also highlight the need for a more systematic and scalable approach to air quality data management, one that reduces reliance on proprietary systems and avoids the need to build bespoke systems from scratch. To that end, a project is underway to build an open-source DMS, informed by potential users. This customizable DMS will be built to handle core functions, such as data ingestion, aggregation, harmonization, storage, quality control, validation, the ability to share data, and more, depending on needs identified by stakeholders around the world. The creation of the DMS will be driven and informed by a global, stakeholder-driven network called COMPASS (Community of Practice for Air Quality Systems).Keywords: DMS, air quality, data analysis.
   
• Air Quality Data Exchange - Standardizing Sensor Data Formats
• Presented by: Ashley Collier-Oxandale, Colorado Department of Public Health and Environment
  
  The Colorado Department of Public Health and Environment has worked with partners in industry and other air agencies to develop a standard format for sharing air monitoring data. The Air Quality Data Exchange (AQDx) format is a groundbreaking solution to address the growing challenges associated with transferring, integrating, and effectively using air quality data from diverse sources and vendors. AQDx offers a way to share not only data, but comprehensive information about how the data was collected and its quality - enabling appropriate use of the data. Furthermore, the format has been designed to meet the needs of a range of data types from sensor-based to regulatory-grade data. This presentation will provide an overview of the latest version of the AQDx format, share examples of how the standard is currently being used, and highlight its suitability for sensor data. In particular, we will discuss the development of supplemental method codes for sensors to complement the existing EPA AQS codes. This presentation will also provide an ideal opportunity to solicit feedback from the air sensor community.
   

Validating Air Sensors and Expanding Data Networks

Session chairs: Achim Haug, AirGradient and Nguyen Thi Kim Oanh, Asian Institute of Technology

• Performance evaluation of ~10 commercially-available and custom-built PM2.5 air sensor systems at an urban core and urban downwind site in a megacity
• Presented by: Daniel Westervelt, Columbia University
  
  The use of air sensors for air pollution monitoring has grown substantially in the past 15 years. A large challenge with air sensors remains their accuracy, but especially for PM2.5 sensors, data quality is rapidly improving. Here, we co-locate about 11 different types of air sensor systems, including those from well-known vendors as well as some custom-assembled systems, with PM2.5 Federal Equivalent Method (FEM) class of monitors at an urban site in New York City (Queens) and an urban downwind/background site in Palisades, New York. Sensors and FEM sampled the same air at both locations for at least 5 months in the summer and autumn seasons in 2023 and 2024, which includes the 2023 Canadian wildfire smoke event, which blanketed New York and the mid-Atlantic states with PM2.5 concentrations reaching 500 µg m-3. We evaluate the performance of these sensors using metrics such as correlation coefficient, normalized mean absolute error (MAE), coefficient of variation, and normalized root mean squared error. Initial results indicate RMSE of sensor performance compared to FEM ranged from 3.9 to 28.6 µg m-3 across both sites and r2 from 0.08 to 0.78. In smoky conditions, RMSE values were 50% worse compared to non-smoky conditions. Sensor performance was poorer at the urban site compared to the downwind site, suggesting that freshly emitted particles may be detected less accurately than aged particles when using optical devices. Sensor types that rely solely on nephelometric methods performed worse overall compared to solutions that leverage optical particle counters and nephelometers. RMSE and r2 were both substantially worse for PM10, with the exception of sensor systems that included optical particle counters, in which PM10 was nearly as accurate as PM2.5.  These results can help improve accuracy and understanding of PM2.5 sensors in urban and downwind New York City.
   
• Evaluation of low-cost sensors in monitoring of surface ozone in different environmental conditions
• Presented by: Kim Oanh Nguyen, Asian Institute of Technology
  
  Tropospheric ozone is toxic to human health, phytotoxic hence can affect crop yields, and is an important GHG. Over the last century, the tropospheric ozone concentrations in the world have increased rapidly due to the increase in precursors' emissions and the warmer atmosphere that enhances photochemical reactions forming ozone. Surface ozone is an emerging pollutant in Southeast Asia (SEA) with high levels observed in large urban areas, but lack of systematic monitoring data hinders a comprehensive assessment of its effects in the region. A Microsensor Challenge project was organized in 2023 to evaluate the performance of low-cost sensors (LCS) in different seasons and environmental conditions, focusing on monitoring fine particulate matter (PM2.5) and ozone air quality. A total of 44 ozone LCS (from 13 brands) were co-located in Lille (France), and 24 ozone LCS (from 8 brands) in Bangkok (Thailand). The data obtained from the LCS were compared with the co-located reference instrument at each site. The results show that the performance of LCS in both places varied widely depending on the type of LCS. For hourly ozone concentrations, the range of Pearson correlation (r) and RMSE was 0.06-0.96 and 9-228 ppb in Lille, respectively, and 0.57-0.97 and 13-93 ppb in Bangkok, respectively. The performances of the LCS were also assessed by considering the range of temperatures, and relative humidity, in conjunction with the levels of potential interfering pollutants such as CO, NO, and NO2.
   
• Performance evaluation of an optical particle counter in Mexico City
• Presented by: Anil Namdeo, Northumbria University and Elizabeth Vega, UNAM Mexico
  
  This study reports on an optical particle counter (OPC) performance evaluation against a reference Beta Attenuation Mass (BAM) monitor in Mexico City.
   
• Evaluation of Sensor Systems Without Co-Location Correction
• Presented by: Kanang Sivula, Vaisala Oyj, Finland
  
  Air pollution is a major global health and environmental challenge, requiring accurate and widespread monitoring to assess its impact and inform policy decisions. Traditional air quality monitoring stations provide highly precise data but are expensive, complex, and often limited in coverage. Low-cost sensors (LCS) offer a promising alternative, enabling more extensive spatial and temporal measurements at a fraction of the cost. These sensors facilitate real-time air pollution monitoring, community engagement, and data-driven interventions, particularly in underserved or remote areas. Co-locating low-cost sensors (LCS) with high-quality reference instruments is widely recognized as essential, as sensor measurements can be influenced by environmental conditions, sensor degradation, and cross-sensitivities to other pollutants. However, this process requires considerable effort. Identifying suitable reference stations—often operated by governmental or research institutions with strict protocols—is the first challenge. Once selected, LCS must be placed near these stations for an extended period, typically weeks to months, to collect comparable data across various environmental conditions. Additionally, extensive post-processing is necessary to apply calibration models and correct for sensor drift over time, ensuring the accuracy and reliability of the measurements. This paper examines the potential of Vaisala's new air quality sensor, the AQT560, by evaluating its performance without co-location-based correction. The Vaisala air quality sensor, manufactured in Finland, undergoes extensive factory calibration to minimize the need for co-location-based correction. This calibration process involves exposing the sensors to controlled laboratory conditions that simulate real-world air pollution scenarios. By using high-precision reference instruments and advanced algorithms, Vaisala ensures that each sensor is pre-calibrated to account for environmental factors such as temperature, humidity, and cross-sensitivities to other pollutants. This approach enhances measurement accuracy and stability across diverse locations, reducing the reliance on traditional co-location with reference stations. As a result, the AQT560 sensor can provide reliable air quality data immediately upon deployment, making it a practical solution for widespread monitoring without the logistical and resource-intensive process of co-location calibration.
   
• High-resolution air quality monitoring of particulate and gaseous pollutants in Bengaluru, India
• Presented by: Emil Varghese, CSTEP
  
  Hybrid air quality monitoring - complementing reference-grade instruments with low-cost sensors (LCS) - is revolutionising air quality monitoring globally. Bengaluru is a metropolitan city in southern India, where a multipollutant (particulate and gaseous) air quality monitoring network comprising 60 low-cost sensors (5-25 nodes each from five different Indian integrators) has been deployed to complement a network of 15 reference-grade air quality monitoring stations (AQMS; 14 government stations and one private station). The sensor network was designed based on satellite data and point-of-interest areas in the city. Before deployment, these sensor nodes were initially collocated for 4-6 weeks with an AQMS at the India Sensor Evaluation and Training (Indi-SET) centre in Bengaluru. Correction models for PM2.5, PM10, NO2, O3 and CO were developed using regression and machine learning methods to improve sensor data accuracy. The sensors used include electrochemical sensors (Alphasense (A4 or B4 series) or EC Sense (TB600 series) for the gases and optical sensors (Plantower PMS7003 or PMS5003, Tera Sense New Gen OEM, Sensirion SPS30 and/or Alphasense OPC-R2 or OPC-N3) for particulate matter (PM). One node from each integrator was held back for a year at the Indi-SET facility to assess the long-term performance of these sensors.  The localised correction of sensors reduced errors to 10-40 %, achieving a correlation with reference instruments greater than 0.7 and ensuring uniform performance across different integrators. Preliminary analysis of the monitoring network indicated that PM2.5 levels at most locations across the city are similar except for some industrial and kerbside locations. Kerbside locations are also hot spots for PM10. NO2 showed spatial and diurnal variations across different nodes, with peaks in the morning and evening traffic rush hours; the north side of the city also showed higher levels of NO2 compared to the south side. Ozone was relatively uniform across the city. The presentation will also cover the long-term performance of multipollutant sensors.
   
• Calibrating Low-Cost Sensors Without Direct Access to Reference Instruments: A Case Study from Pai, Thailand
• Presented by: Anika Krause, AirGradient Ltd
  
  Resource-limited regions frequently suffer from high levels of air pollution, yet they often lack the infrastructure needed to effectively monitor and mitigate its impacts. Low-cost sensors offer a potential solution, but concerns about their accuracy and reliability persist. Traditional calibration methods rely on co-locating sensors with local reference instruments to extract calibration parameters tailored to local conditions. However, reference instruments are often inaccessible due to their high cost and limited spatial distribution in many regions.In this study, we demonstrate how sensor networks can integrate public reference data from the wider region to improve calibration without direct access to reference instruments. In January 2024, a network of 16 AirGradient monitors was deployed in Pai, Thailand, at key locations such as a health center, a border patrol station, the mayor’s residence, and the immigration office. This network enabled real-time air quality monitoring, particularly during the burning season when pollution levels peak.By extracting background PM2.5 levels from the network and comparing them with publicly available data from a reference station 50 km away in Mae Hong Son, we derived calibration parameters that significantly improved sensor accuracy. Additionally, the network allowed for continuous monitoring of data quality by detecting sensor drifts and malfunctions through comparisons among sensors.These methods provide a scalable, low-cost solution for improving air quality sensor data, enabling more effective pollution monitoring and mitigation in resource-constrained regions.
   
• Solving the difficulty of calibrating NO chemical low-cost sensors for air quality monitoring
• Presented by: Michel Gerboles, European Commission, Joint Research Centre
  
  The field calibration of NO chemical sensors typically involves establishing a multivariate equation to adjust for the effects of influencing parameters, utilizing co-located sensors and reference method data. The aim of this equation is to correct for the strong influence of temperature (T) and humidity (RH) to NO sensor responses which shows an exponential relationship. The extent of T and RH influence can be significant in particular when the NO concentrations are low accompanied by large T and RH changes.The laboratory experiments conducted under controlled conditions of T and RH at varying NO concentrations revealed the effects of these parameters on NO sensor response. For both increasing and decreasing T, the effect shows a reproducible exponential growth. In contrast, for increasing RH, the sensor response shows exponential growth, while an exponential decay is observed for decreasing RH. Therefore, at a field site, one can expect that the effect of the daily cycle of RH changes is incorporated into a hysteresis cycle of sensor response, which depends on the history of RH changes and leads to drifts that are challenging to predict. A field calibration method based on a multivariate correction for T and RH without accounting for this RH hysteresis pattern have limited validity, with significant drift inevitably appearing over time.There are substantial evidences in the literature indicating the poor performance of NO sensors compared to reference measurements when NO sensor predictions are determined based only on one multivariate calibration for a long predicting period.We propose a new strategy for predicting NO concentrations based on baseline correction of NO sensor response. This baseline correction involves identifying all time-regions exhibiting changing trends in both T and RH on a daily basis. In each region, an exponential model accounting for both T and RH effects is fitted according to the laboratory results. This approach ensures that all sensor predicted NO remain drift-free. NO levels are determined by dividing the baseline-corrected sensor responses by the sensor's sensitivity in nA/ppm, which can be established through laboratory tests and appears to remain reasonably constant over one year. This approach does not require an initial field calibration.Finally, the performance of the baseline correction approach is compared to the standard method with one equation of multivariate correction using the results of a pilot study involving 85 sensor systems deployed in 3 European cities for a year.
   
• Optimizing Urban Air Quality Monitoring Networks: A Multi-Method Approach for Spatial-Temporal Efficiency and Cost-Effective PM2.5 Surveillance in Kampala, Uganda
• Presented by: Fidel Raja Wabinyai, AirQo
  
  Urban air quality monitoring is critical for public health, particularly in rapidly growing cities like Kampala, Uganda, where pollution levels rise alongside urbanization. The AirQo network, deploying over 60 low-cost PM2.5  sensors, provides vital air quality data but faces challenges in balancing spatial coverage, temporal reliability, and operational costs. This study proposes a multi-method optimization framework to enhance the network’s efficiency. First, spatial distribution analysis employs spatial autocorrelation (Moran’s I, Getis-Ord Gi*) and Voronoi diagrams to evaluate coverage gaps and redundancies. The second, temporal analysis investigates diurnal, seasonal PM2.5 trends and device redundancy via cross-correlation and downtime assessments. Third, machine learning and clustering (K-means, feature importance, simulated annealing) optimize sensor placement to minimize redundancy while preserving data accuracy. Finally, cost-benefit analysis quantifies financial savings and data quality trade-offs, incorporating stakeholder feedback to ensure policy relevance. This research aims to deliver actionable strategies for cost-effective, high-fidelity air quality monitoring by integrating spatial, temporal, computational, and economic perspectives. The findings will guide policymakers in Kampala and similar cities to prioritize sensor deployment, reduce operational costs, and strengthen evidence-based environmental management. This holistic approach underscores the value of interdisciplinary methodologies in addressing urban sustainability challenges.Keywords: Air quality monitoring, PM2.5 , sensor optimization, spatial-temporal analysis, machine learning, cost-benefit analysis.
   
• A fusion approach at estimating PM 2.5 in Laos – coupling low cost sensors and open source satellite imagery
• Presented by: Hugo Ruiz Verastegui, UNICEF
   
• Expanding air quality monitoring networks in Metro Manila through sensors: from data correction to AQM action
• Presented by: Everlyn Tamayo, Clean Air Asia
  
  Megacities are at the highest risk of air pollution health impacts, with high population densities exposed to concentrated sources of pollutants. In Asia, all megacities except for Tokyo had annual PM2.5 in 2022 which exceed the WHO Air Quality Guidelines. The National Capital Region (NCR or Metro Manila) of the Philippines is composed of 16 cities and 1 municipality, with a total population of 14.4 million. The national and local government agencies aim for the expansion of the air quality monitoring network in the region, and air sensors provide an opportunity to fill data gaps, engage stakeholders, and inform air quality management (AQM).An important step in the use of non-reference air monitors is the evaluation of sensor performance, ensuring that sensor measurements are accurate, reliable, and as close as possible to reference-grade monitoring instruments. The use of uncorrected raw sensor data can lead to misinformed actions, unnecessary panic, unfounded complacency, and public confusion. Clean Air Asia highlights this requirement as we support various cities in Metro Manila in the establishment and expansion of their air quality monitoring networks, bridging the gap towards the accessibility of accurate air quality data for action planning and implementation.We present the results of the collocation studies of several units from three (3) sensor brands with the Environmental Management Bureau’s reference monitoring stations, accomplished as part of the 3M Asia Blue Skies Program and the Quezon City Air Quality Management Project. This study provides insight on the necessary inclusion of temperature and humidity to correct raw sensor data through multiple linear regression. Aside from determining correction coefficients, variations in the mean average error (MAE), root mean square error (RMSE), and relative accuracy are explored to determine sensor performance across time and in various PM2.5 concentration ranges. The results are envisioned to provide guidance in the development of the final sensor evaluation guidelines in the country and to help government, academia, and civil society stakeholders in understanding sensor data. Lessons learned on the logistical, manpower, and technical needs related to sensor evaluation and deployment will also be shared. Overall, the correction of sensor data based on the collocation performed increased the confidence of the government and stakeholders on the information from the sensor network, and strengthened the engagement of the public.
   

Communication and Public Engagement Using Air Sensors

Session chairs: Jim McQuaid, University of Leeds and Kayla Schulte, Imperial College

• Assessment of Outdoor Workers' Air Pollution Exposure in Central Bangkok Using Portable Air Sensors
• Presented by: Nutta Taneepanichskul, Chulalongkorn University
  
   
• Engaging High School Students in Air Quality Awareness: The "Air Pollution Detective" STEM Activity
• Presented by: Eliani Ezani, Universiti Putra Malaysia (UPM)
  
  Public engagement in air quality science is essential for fostering awareness and advocacy, particularly among younger generations. As part of a STEM Day event, we designed and conducted an interactive activity called "Air Pollution Detective" at a high school in an urban area of Malaysia, involving 86 students aged 13 to 16. The activity aimed to introduce students to air pollution concepts through hands-on experience with low-cost air sensors. A dedicated station was set up where students were briefed for 10 minutes on air pollution and the use of low-cost sensors. Two Plume Labs Flow units were provided, with facilitators guiding participants in tracking air quality around the school compound within a 30-minute timeframe. The students recorded air pollution levels and, at the end of the session, mapped out safe walking or commuting routes based on their findings. Post-activity feedback was collected through a survey (n=86), revealing that 93% of participants found the activity engaging and informative, sparking their interest in air pollution science. However, 70% felt that the allocated time was insufficient. Despite the limitations, this initiative demonstrated that low-cost sensors, while not research-grade, serve as effective tools for introducing students to air quality measurement. The "Air Pollution Detective" activity highlights the potential of integrating hands-on environmental monitoring into school programs to enhance STEM education and promote public awareness of urban air pollution.
   
• Empowering Communities with Open Air Quality Data in Pakistan: Lessons from Eight Years of Sensor Deployment
• Presented by: Abid Omar, Pakistan Air Quality Initiative
  
  Pakistan faces some of the world’s highest air pollution levels, yet much of the country remains a ‘data gap,’ with limited monitoring infrastructure and minimal scientific research. The Pakistan Air Quality Initiative (PAQI) has worked to close this gap by building a nationwide community-driven air sensor network that provides open, real-time data to inform public awareness and advocacy.Over eight years, PAQI’s sensor network has delivered critical insights, identifying pollution hotspots, informing public discourse, and empowering citizens to demand action. This presentation will highlight key case studies where data from low-cost sensors influenced media coverage, drove policy conversations, and engaged underserved communities in air quality awareness.The presentation will also address the technical and operational challenges of deploying sensors in resource-constrained environments, including power disruptions, connectivity issues, and sensor maintenance. By sharing these experiences, PAQI aims to inspire new partnerships and collaborations to expand sensor networks and promote open-access data in regions facing similar challenges. PAQI’s work demonstrates how low-cost sensor networks, combined with effective communication strategies, can empower communities and drive meaningful change in air quality awareness and policy.
   
• Framing air pollution as an inequity issue to increase public engagement
• Presented by: Tanushree Ganguly, University of Chicago
  
  To drive public and policy engagement on air pollution, various framings have been used to highlight its impacts—ranging from health and economic costs to environmental and social justice concerns. However, the unequal burden of air pollution in low- and middle-income countries remains understudied due to limitations in timely, socio-economic, and spatially representative pollution data.This study addresses this gap by examining global and within-country inequalities in air pollution’s impact on life expectancy. Given the scarcity of ground-based air quality data in many highly polluted regions, we use satellite-derived estimates of fine particulate matter (PM2.5) concentrations at regional, national, and global scales. Applying the Air Quality Life Index (AQLI) (Greenstone et al., 2018) to these estimates, we quantify life expectancy losses due to air pollution and assess inequality by comparing the most and least polluted quintiles worldwide and within individual countries.Our analysis reveals stark disparities. Globally, people living in the most polluted regions (top quintile) lose, on average, three more years of life expectancy compared to those in the cleanest regions (bottom quintile). In the ten most polluted countries, residents in the most polluted areas could gain up to 2.4 additional years of life expectancy, compared to those in the least polluted areas.While these findings expose significant inequalities at the country level, they do not capture hyper-local disparities in exposure, given the coarse resolution of satellite-derived pollution data. Community-driven, localized assessments could help build a stronger case for targeted clean air interventions. This presents a unique opportunity to promote citizen science by engaging communities in air quality monitoring using sensor-based technologies.As part of AQLI’s efforts to advance local air quality action, we aim to collaborate with groups generating hyper-local pollution data and support them in communicating the health benefits of reducing exposure disparities through localized clean air interventions.
   

Community-Driven Air Quality Monitoring and Equity

Session chairs: Garima Raheja, Columbia University and Saumya Singh, Indian Institute of Forest Management Bhopal, India

• Community-led, crowd-sourced odor reporting and air quality monitoring in Pittsburgh, PA as a model
• Presented by: Mike Tasota, CREATE Lab
  
  The CREATE Lab at Carnegie Mellon University is both a technology development ground and a community partner. It is this unique combination that enables a new form of local change: one that empowers people to chart their technology future and, most important of all, their community's prospects for quality of life. Air pollution causes serious health impacts that disproportionately affect underserved communities in our Appalachian region. Our work supports our region’s residents in their efforts to demand better, and create greater awareness of this health and environmental justice issue across all stakeholders in our region. Awareness and participation are important precursors to the movements toward equity. For 10+ years, CREATE Lab has worked with partners to co-design tools to document local pollution. This work has led to the development of technical and sociological practices that have genuinely empowered communities to understand and elucidate their relationship with local pollution, and work to improve our air quality. We use SmellPGH/SmellMyCity apps, BreatheCams, and our Environmental Sensing Data Repository to build on the combined power of crowd-sourced reporting, continuous sensor measurements and spectroscopy, and time-lapse imagery and visualization to engage, empower, and inform the public and policy makers. 
   
• Strengthening Community-Driven Air Quality Monitoring through Participatory Budgeting and Inclusive Urban Governance: A Case Study of Breathe Bangkok Initiative
• Presented by: Pongpisit Huyakorn, Chulalongkorn University
  
  Air pollution remains a critical urban challenge, disproportionately impacting vulnerable communities with limited access to real-time air quality data and decision-making processes. While low-cost air sensors have made air quality monitoring more accessible, the lack of community involvement in governance and resource allocation has restricted their full potential. Furthermore, despite efforts to map air pollution levels and their health impacts, there remains a significant gap in understanding how air pollution, public health, and inequality intersect. The absence of a centralized system integrating air quality data with public health and socio-economic indicators limits the ability to develop targeted interventions. This paper explores how participatory budgeting (PB) and inclusive urban clean air governance—leveraging localized air quality, GHG emissions and health data —can empower local communities in shaping air quality policies and interventions.As part of the Breathe Bangkok initiative, this project will integrate low-cost air sensors with a PB framework across 15 pilot districts in Bangkok, enabling residents to propose, prioritize, and implement air quality interventions based on localized data. The project will work in collaboration with the Bangkok Metropolitan Administration (BMA), district offices, and public health centers, engaging local health volunteers to audit existing government clean air initiatives, assess air and health quality indexes, and develop participatory budgeting proposals to secure funding from BMA and National Health Security Office (NHSO). By embedding community-led air quality and health monitoring within municipal budgeting structures, the project seeks to promote transparency, accountability, and more equitable allocation of resources and sustainability of city's clean air initiatives, ensuring that underrepresented groups—including informal workers, low-income households, and the elderly—play an active role in environmental decision-making.This study will focus on three core areas: (1) Data Democratization and Governance, ensuring communities access air quality data for advocacy and self-protection; (2) Participatory Budgeting as a governance tool, directing municipal funds to community-led air quality initiatives; and (3) Policy Uptake and Institutionalization, examining how community-generated data and PB outcomes influence long-term regulatory frameworks.The project will assess how integrating PB with community air monitoring can drive targeted,locally tailored interventions that are more relevant, inclusive, and actionable, while addressingchallenges such as city-wide scalability and bureaucratic inefficiencies in governance. It will explore data-driven strategies to overcome challenges such as scaling participatory budgeting city-wide   and   streamlining   decision-making   processes   between   local   governments and communities.By outlining planned activities, expected outcomes, and policy implications, this paper contributes to the ongoing discourse on community-driven environmental governance. This initiative aims to establish participatory budgeting as a scalable and institutionalized model for community-driven air quality governance, ensuring that marginalized groups have both the data and the decision-making power to advocate for cleaner, healthier urban environments.
   
• Advancing equity through air quality monitoring: opportunities for inclusive urban policy
• Presented by: Pongsatorn Greigarn, C40 Cities Climate Leadership, Inc.
  
  Air pollution is more than an environmental challenge—it is a pressing equity issue that disproportionately affects marginalized and low-income communities. While air pollution monitoring and mitigation efforts have increased, significant gaps remain in data accessibility, community engagement, and policy integration. Many cities lack hyperlocal, real-time air quality data, particularly of underserved, low-income areas near waste burning sites, industrial or major transit routes where pollution exposure is highest and residents have fewer resources to protect themselves. Reliance on centralized monitoring systems could overlook neighborhood-level disparities, leaving vulnerable populations unprotected. Additionally, there is a disconnect between air quality data and policy action, often due to unsupported community involvement and governance structures that fail to prioritize equity in clean air governance.  This presentation will share key findings from the Advancing Equity Through Clean Air Action report (set to be released at the World Health Organization’s second Global Conference on Air Pollution and Health in March 2025). It will examine challenges and opportunities in advancing equitable air quality monitoring, emphasizing how localized, hyperlocal air quality monitoring can enhance policy precision and ensure interventions benefit those disproportionately affected. Case studies from Jakarta, Bangkok, and Quezon City will be spotlighted to show how inclusive air quality policies—such as community-led monitoring networks and targeted pollution reduction initiatives—can drive environmental justice. The presentation will conclude with practical recommendations for policymakers, urban planners, and civil society to drive inclusive, equitable clean air action across the region. 
   
• PM2.5 Monitoring with Low-Cost Sensors and the Development of Community Tools for Air Quality Measurement in the Thai-Lao Border Area of Northern Thailand
• Presented by: Wan Wiriya, Chiang Mai University
  
  This study aims to enhance air quality monitoring and community engagement using low-cost sensor technology and digital tools, while also assessing fine particulate less than 2.5 micrometer (PM2.5) variations in this transboundary region. This research deployed DustBoy PM2.5 sensors at 16 monitoring sites in Nan Province, Thailand, and 3 sites in Xayaburi, Laos. These low-cost sensors provide real-time air quality data, offering a more localized and frequent alternative to traditional air quality monitoring stations. The data collected helps identify PM2.5 pollution trends, particularly during the dry season when agricultural burning is prevalent. Our findings show significant seasonal fluctuations, with PM2.5 concentrations peaking between February and April, aligning with biomass burning activities in the region. During this period, PM2.5 levels in Laos were consistently higher than in Thailand’s border areas, corresponding to a higher number of fire hotspots detected in Laos.  Recognizing the need for accessible and user-friendly air quality information, we integrated the Chiang Mai University (CMU) LINE Official account as a digital engagement tool. This platform allows local residents, policymakers, and researchers to track real-time PM2.5 levels, receive alerts, and access health advisories directly on their mobile devices. This study highlights the potential of low-cost sensor technology and community-driven air quality monitoring in addressing air pollution challenges in border regions. By bridging data gaps and fostering local participation, this approach provides a scalable and cost-effective model for enhancing public awareness, supporting evidence-based policy decisions, and strengthening transboundary air quality management. Moving forward, expanding this network and integrating additional digital tools could further improve the effectiveness of air quality interventions in Southeast Asia.
   
• Evaluating the impact of real-time, public air quality informational displays: unpacking the AWAIR Project
• Presented by: Kayla Schulte, Imperial College
  
  Air pollution remains a critical public health challenge, especially in urban areas, where residents face high levels of harmful pollutants. Despite this, many communities lack access to real-time information that could help mitigate air pollution exposures. The AWAIR project addresses this gap by providing co-created, accessible air quality information in real-time via public displays in three neighborhoods across London, UK. AWAIR aims to support and work with communities by offering air pollution information that can reduce personal exposure and stimulate community-driven efforts to lower emissions. The project also explores how real-time information can inform broader public health strategies and guide future interventions.The methodological approach for this project involved identifying three London neighborhoods or ‘wards’ for the pilot display deployments using social and environmental datasets, while working alongside local community groups. A baseline survey was conducted to evaluate residents’ existing knowledge, attitudes, and behaviours related to air pollution. The survey also examines how sociodemographic factors such as age, gender, income, and education correlate with awareness levels. This was followed up by co-design workshops with residents to ensure the displays were relevant, accessible, and culturally appropriate. The workshops led to the creation of a new air quality communication scale and messaging that incorporates both the World Health Organization (WHO) guidelines and the UK’s Daily Air Quality Index (DAQI).A second survey was then conducted to evaluate changes in knowledge and behaviour. Logistic regression analysis is used to identify key social and environmental factors associated with display interaction and uptake of exposure reducing behavioural recommendations delivered through the display. Comparative analysis with national datasets supports our evaluation of whether local engagement patterns with air pollution information align with national trends.Overall, this study provides valuable insights into the efficacy of community-guided air quality displays in raising public awareness and supporting exposure reducing behaviour. By comparing local and national trends, the project aims to inform future interventions, equipping policymakers and public health officials with the tools to better tailor strategies to different demographic groups. Ultimately, this research contributes to the broader goal of reducing public exposure to harmful air pollutants and improving health outcomes at the local level.
   
• Beyond Megacities: Revealing Rural Air Quality Disparities in the Indo-Gangetic Plain through Community-Based Monitoring
• Presented by: Mark Campmier, UC Berkeley
  
  High emissions from a diverse source mixture and unfavorable meteorology have resulted in extreme fine particulate matter (PM2.5) air pollution in the Indo-Gangetic Plain (IGP) of North India. Most studies and abatement strategies have focused on densely populated megacities, but the majority of the population in the region (60-70%) resides in non-urban settlements. Therefore, although non-urban air quality is understudied in the IGP, it is critical to understand the regional health burden of PM2.5 and develop informed policy. We implemented a lower-cost PM2.5 monitoring network of over 80 sites across more than 15 settlements using the popular PurpleAir sensor. Our network spanned megacities to regional background sites and represents the first ground-based observations of PM2.5 in most settlements. We observed sustained poor ambient air quality across the region (annual average PM2.5 concentration ≥ 60 μg m−3), with weak spatial gradients from urban core sites to regional background sites. Leveraging the high temporal resolution afforded by lower-cost sensors, we observed that although annual and seasonal trends were strongly correlated, diurnal patterns diverged across settlement population density strata. Megacities featured smooth diurnal profiles, with peaks later in the evening and larger intra-settlement variability, indicating the complex mixture of traffic, heavy industry, and other sources. Conversely, non-urban and small city sites featured earlier and higher magnitude diurnal peaks (1.5 -1.7 x daily minimum), with low intra-settlement variability, likely indicating a higher relative prevalence of biomass burning. Policymakers should ensure clean air programs do not solely focus on relatively well-studied megacities at the expense of the large non-urban population distributed across the IGP.
   
• Community-based environmental pollution monitoring built on pervasive IoT sensor network of air quality monitors
• Presented by: Eric Morris, A.U.G. Signals Ltd
  
  There is limited quantitative research on how and why members of public engage with air quality sensors and data, with most studies involving small participants groups of 10-20. This study presents ongoing field trials conducted near school campuses close to roadways in & around the city of Cork, Ireland. The trails aim to explore two key aspects: 1) how an IoT sensor network of air quality monitors, such as AirSENCE, can shape individuals and community perceptions of air quality information; and 2) how this information can be leveraged to improve standards of living.The trials commenced with the deployment of a network of AirSENCE devices in 2022 to collect and provide real-time air quality data. A series of expert interviews/surveys are being carried out with potential stakeholders—including city and county councils, planning authorities, and industries—to identify the boosters and blockers to making use of air quality data. Qualitative research then focuses on end-users, namely secondary school/college students, parents, and some pollution generators (e.g., trucking companies/drivers), who keep daily diaries of activities in the context of their pollution exposure and creation. The trials aim to verify that access to enhanced air quality data and analysis can help the larger population of the city to reduce pollution through actions such as promoting car-sharing among parents and encouraging local hauliers operating on nearby roads to reschedule travel outside school drop-off and pick-up times.
   

Posters & Lightning Talks

Lightning talks indicated by *.

• On the origin of particles - a method combining low-cost particle monitors with filter sampling and spectrographic analysis to trace the source of particles*
• Presented by: Paul Baynham, Mote Limited
  
  To effectively manage ambient particulate matter (PM) in high pollution areas, it is essential to understand the source and quantity of PM emissions. In complex airsheds with multiple PM sources, quantifying ambient contributions of PM from specific sources can be challenging and resource-intensive. Traditional techniques such as dispersion modelling or emission inventories often struggle to accurately determine PM source contributions during high pollution events due to short-term variability in emissions.This method employs low-cost particle sensors combined with smart filter sampling, followed by automated scanning electron microscopy (SEM) and spectrographic analysis to identify contributions from natural and specific emission sources. The ambient PM sampling instruments utilize solar-powered systems, facilitating remote deployment to ensure adequate spatial coverage. The initial SEM analysis may be supplemented with one or more spectrographic techniques, including Raman spectroscopy, X-ray photoelectron spectroscopy (XPS), atomic force microscopy-infrared (AFM-IR), or micro-Fourier transform infrared spectroscopy (Micro-FTIR), where a higher level of confidence is required to associate an emission source with ambient particulate.Two case studies from New Zealand demonstrate this approach: one at an industrial port and another assessing emissions from a commercial poultry farm affecting nearby properties where this methodology was used to successfully attribute particles to specific sources and activities.
   
• Ozone Gas-Sensitive Semiconductor Sensor Validation Study in Indonesia*
• Presented by: Driejana Driejana, Institut Teknologi Bandung, Indonesia
  
  Ozone is a secondary pollutant produced by photolysis reactions of nitrogen dioxides. In the Northern Hemisphere, ozone problems generally are found in the summer.  The Indonesian tropical region receives abundant solar radiation throughout the year, so high ozone concentration in urban areas happens almost daily. Ozone may cause short-term impacts, e.g., acute asthmatic disorders, respiratory tract irritation, and heart attacks. The prolonged ozone pollution could lead to chronic health problems in the form of permanent tissue damage to the respiratory system. The oxidative properties of ozone might also reduce rice production and damage to vegetables and fruits.Monitoring ozone concentrations on a real-time basis is needed to prevent health impacts. Ozone is one of the regulated air pollutants and is continuously monitored by the Government using FEM-grade analyzers in a limited number of locations.  Low-cost sensors (LCS) might have the potential to provide better spatial coverage.  This study aims to validate the performance of gas-semiconductor sensors (GSS) in the Indonesian environment. We co-located two units of Aeroqual S500 GSS  with an APOA-370 NDUV Absorption Ozone Analyzer of the Jakarta Province Environment Agency (DLH) in a monitoring location in South Jakarta.Since we are based in Bandung, to be able to access data remotely, S500 sensors were equipped with an IoT system through the analog output. We deployed the Aeroqual S500  for 53 days in April - June 2022. We set Aeroqual S500 to report data every 2 minutes. We then average the sensor data to a 30-minute average for comparison with the DLH analyzer data. We gathered meteorological measurements data of air temperature, RH, and pressure from DLH monitoring. We found S500 precision of 80.147%. Comparison with APOA-370 resulted in r2 = 88.45% (p = 0.000). During the measurement, the average temperature and RH ranged between  23.66 - 35.92 °C and 37.88 - 99.93 %.  The performance of S500 is considered good after continuous field exposure of more than 30 days. The GSS S500 produced reasonably accurate data comparable to the FEM analyzer. The GSS can be used in campaign measurement in various air quality management studies. Sensor lifetime for continuous measurement still needs further investigation.
   
• Assessing PM2.5 Exposure of Active Commuters Using Mobile Monitoring on a Simulated Walking Route in Kuala Lumpur*
• Presented by: Eliani Ezani, Universiti Putra Malaysia (UPM)
  
  Personal exposure to traffic-related air pollution presents substantial health risks, especially in densely populated urban areas. This study aims to assess the exposure of active commuters to PM₂.₅ near roadside environments using a research-grade instrument (SidePak TSI) and compare its performance with a low-cost sensor (Atmotube). Both sensors were mounted on a backpack and deployed to measure PM₂.₅ exposure along pedestrian kerbside during three commuting rush-hour periods in Kuala Lumpur. Statistical analysis demonstrated a moderate-to-strong correlation between the two sensors (R = 0.40); however, the SidePak consistently recorded higher PM₂.₅ values, likely due to its superior sensitivity and response to high-concentration roadside pollution. In contrast, the Atmotube exhibited smoother or lower PM₂.₅ variations, potentially due to sensor lag or its averaging algorithm and proned to underestimate PM₂.₅ levels compared to the SidePak, particularly at higher concentrations. PM₂.₅ concentrations observed peak levels during morning and evening commutes. An inhalation dose assessment was conducted, incorporating factors such as walking speed, breathing rate, and pollutant concentration. The results indicated that the estimated daily inhaled dose of PM₂.₅ for an average commuter was highest during the morning rush hour, coinciding with peak traffic emissions and reduced atmospheric dispersion. This study highlights the importance of applying calibration and correction factors when utilizing low-cost sensors for public health evaluations and offers critical insights into pedestrian exposure risks in Southeast Asia.
   
• Deployment and Preliminary Assessment of PM2.5 and NO₂ Measurements by Clarity Node-S in Suburban and Industrial Areas of Malaysia*
• Presented by: Eliani Ezani, Universiti Putra Malaysia (UPM)
  
  Particulate matter (PM₂.₅) and nitrogen dioxide (NO₂) pollution are major contributors to global environmental health issues, with Southeast Asia being especially affected due to rapid industrialization, economic growth and heavy reliance on private vehicles. This preliminary study examines the deployment of demo unit of Clarity Node-S sensors at two locations in Malaysia: a suburban university campus in the Klang Valley and an industrial zone in Pasir Gudang, Johor. The aim was to compare PM₂.₅ and NO₂ concentrations at these sites and assess the sensors’ performance under different environmental conditions. The sensors operated continuously from May to September 2024. Our campaign observed distinct pollution patterns between the two areas, with higher levels of NO₂ and PM₂.₅ in the industrial zone, likely due to emissions from factories and heavy traffic. In contrast, the suburban site exhibited lower concentrations, though occasional spikes were observed, possibly linked to surrounding activities. Sensor readings were compared with available reference data to evaluate accuracy and reliability. Findings from this study provide insights into the effectiveness of low-cost sensors for air quality monitoring in tropical environments. The results highlight their potential for expanding air pollution monitoring, particularly in areas lacking regulatory monitoring networks.
   
• Developing Bangkok’s First Clean Air Management Plan: A Roadmap for Sustainable Air Quality Improvement
• Presented by: Pongsatorn Greigarn, C40 Cities Climate Leadership, Inc.
  
  Effective air quality management requires a comprehensive monitoring framework and a clear policy roadmap to drive sustainable improvements. In Bangkok, the Breathe Cities initiative is supporting the development of its first-ever long-term Clean Air Management Plan to reduce PM2.5 levels and align with Thailand’s national air quality standards and World Health Organization Air Quality Guidelines. This multi-year plan will integrate air quality monitoring, emissions controls, stakeholder engagement to transition Bangkok towards sustainable clean air management. It will align with the forthcoming national Clean Air Management Act, incorporate regional best practices, and strengthen institutional capacity for effective policy implementation. The Clean Air Management Plan will also mainstream sectoral policies, prioritize public health outcomes, and promote inclusive governance, ensuring participation from vulnerable communities and key stakeholders. To sustain long-term air quality improvements, the plan will enhance monitoring, evaluation, and enforcement mechanisms while leveraging Bangkok’s strong policy commitment and dedicated funding for air quality infrastructure. By expanding real-time monitoring networks, including low-cost sensors, Bangkok aims to bridge data gaps, inform targeted interventions, and improve  public health outcomes. This presentation will explore key components of Bangkok’s approach to develop the Clean Air Management Plan, highlighting the intersection of air quality monitoring and policy development. It will also emphasize the importance of data-driven decision-making, inclusive governance and strategic interventions to create a scalable model for effective air quality management in urban environments.
   
• Monitoring Crop Residue Burning: Identifying Partially and Completely Burnt Fields through Remote Sensing*
• Presented by: Rishikesh P., Council on Energy, Environment and Water
   
  Crop Residue Burning (CRB) is a global concern, it is estimated that the world burnt around 458 million tonnes of crop residue in 2019 (FAO, 2022). The particulate matter from crop residue burning worsens air quality in South and Southeast Asian countries during the post-harvest months. Countries such as Thailand, Indonesia, and the Philippines burn more than 45 per cent of rice straw generated (Reddington C., 2022). In 2016, around 24 per cent of rice straw generated was burnt in India (TERI, 2019). Farmers set fire to the residue to clear their fields and prepare them for the next crop due to the short harvest window. Over the years, with improved mechanisation, some farmers have shifted from completely burning their fields to partially burning them (Kemanth et al, 2024). The primary reason for this practice is to reduce the operational costs of managing the residue with machines.
  Satellite-borne sensors can detect active fires and also the burn scars that remain after agricultural burning. This information is then used to estimate the emissions. Both methods have advantages and disadvantages of their own (Earth Data Forum NASA, NOAA). While satellite imagery can detect fully burnt fields reasonably well (Deshpande et. al, 2022), there have been no studies on their effectiveness in detecting partially burnt fields. Applying thresholds meant for fully burnt fields would lead to an underestimation of the burnt area due to the omission of the partially burnt fields and hence emissions. Therefore, in this study, we attempt to devise a method to identify partially and fully burnt fields from satellite imagery in Punjab, India. The method can be applied universally and can improve estimates of greenhouse gases (GHGs) and air pollutants from agricultural burning.
   
• University Clean Air Network (UCAN): A University-Led, Community-Driven Approach to Air Quality Monitoring in South Asia*
• Presented by: Vamsi Krishna, Aurassure Pvt Ltd.
  
  Air pollution is a major concern for public health and environmental crisis in South Asia, contributing to respiratory diseases, cardiovascular disorders, and reduced life expectancy. Despite its severe impact, air quality research remains limited due to the high cost of traditional monitoring devices, leaving critical data gaps that hinder effective policymaking and mitigation strategies. Aurassure’s unique University Clean Air Network (UCAN) addresses this challenge by establishing a university-led, community-driven air monitoring ecosystem. Launched with over 15 universities and schools in India, UCAN has state-of-the-art IoT-based air quality sensors, providing real-time air pollution and meteorology data and enhancing research and community engagement. UCAN’s unique model leverages universities as scientific hubs, bridging academia and policymakers. UCAN enhances environmental equity by sharing data among partner universities and integrating research, technology, and participation. In the long run, UCAN aims to build a sustainable air quality data repository, support policy decisions, and foster regional collaboration to address transboundary pollution challenges.
   
• Black Carbon Monitoring Using Low Footprint Smart Air Quality Stations in Bangkok, Thailand*
• Presented by: Jost V. Lavric, Acoem GmbH, Germany
   
  According to the United Nations Department of Economic and Social Affairs (2018), the global urban population is projected to rise from 55% in 2018 to 68% by 2050, with approximately 90% of this increase occurring in Asia and Africa. The Bangkok Metropolitan Administration (BMA) has been running an air quality network (AQN) for more than 25 years, monitoring a range of critical pollutants at fixed and mobile monitoring locations, covering all the city’s districts (https://airquality.airbkk.com/). Based on six key pollution parameters (PM2.5, PM10, O3, CO, NOx, and SO2), the BMA is providing a spatially resolved near-real time air quality index (AQI). This empowers citizens to make informed decisions about their daily activities to reduce their exposure to harmful pollutants.
   
• Real-Time Hyper-local air quality monitoring for Just in Time Interventions: A pilot study*
• Presented by: Sneha Mahalingam, National Aerosol Facility, IIT Kanpur
   
  Challenges to urban sustainability due to air pollution arises the need for high-resolution and real-time air quality monitoring data. However, the availability of hyperlocal air quality data is restricted due to inadequate monitoring infrastructure and high costs. High-resolution data demonstrates that air pollution levels can fluctuate significantly over short distances, with variations of up to eightfold within a single city block. This necessitates high-resolution spatiotemporal mapping of pollution spikes to accurately identify emission sources in real time within minutes. This research proposes a Just-in-Time (JIT) intervention framework, integrating a suite of technologies which includes (i) network of static portable air quality monitors (AQM) for fine particulate matter (PM2.5), volatile organic compounds (VOCs), and meteorology; (ii) mobile real-time source apportionment laboratory equipped with High-Resolution Time-of-Flight Aerosol Mass Spectrometer (HR-ToF-AMS), Xact-625i, and E-BAM for measuring organics, inorganics, trace elements, PM10, PM2.5, and meteorology, and (iii) drone-based vertical profiling with portable AQM for PM2.5, VOCs, and meteorology. Incorporating low-cost AQM in this approach will significantly reduce expenses compared to traditional monitoring systems. This integrated multi-platform approach (static, mobile real-time source apportionment laboratory, and drone-based air quality monitoring approach) enhances the spatial and temporal resolution of air pollution assessments, enabling the real-time detection of pollution spikes and precise identification of emission sources at high resolution. Based on this, a pilot study was conducted at IIT Kanpur to assess the feasibility of this framework in capturing transient air pollution events from the controlled burning of cow dung cakes, solid waste, and dry leaves. Cross-platform validation demonstrated a strong correlation across with drone mounted, static and mobile-laboratory mounted AQM data, confirming the reliability of portable AQM data from drone mounted and static real-time air quality monitoring. Results indicate that solid waste burning exhibited the highest PM2.5 concentrations (325.8 ± 57.4 μg m-³), while cow dung (221.58 ± 45.63 μg m-³) and dry leaves (98.8 ± 36.3 μg m-³) resulted in moderate but prolonged emissions. The mobile laboratory provided high-resolution spatial mapping of pollution plumes, confirming localized concentration peaks near the burning zones. The combustion of cow-dung cake resulted in the emission of metals (Pb, S, Si), inorganics (SO₄, NH₄, NO₃, Cl), and organic compounds. Solid waste burning released metals (Cl, Pb, S) along with inorganic (NH₄, Cl⁻) and organic constituents. Similarly, the combustion of wood and dry leaves emitted metals (S, Si) and inorganics (SO₄, NH₄), along with organic species. These emission sources were identified JIT after a few minutes of the experiment. Also, drone-mounted measurements revealed a rapid vertical decay in PM2.5 and VOCs with increasing altitude. The JIT intervention framework offers a scalable and effective approach for real-time air quality management, enabling rapid hotspot detection and identification of pollution sources.
   
• Evaluating Machine Learning Algorithms for Improving Network Sensor Data for Air Quality Monitoring*
• Presented by: Julius Mendoza, University of the Philippines, Diliman
  
  Low-cost air quality sensors, while affordable, often exhibit limited precision and require frequent calibration. A previous study by the UP CARE research team investigated the performance of two SPS30 PM2.5 sensors over a 6-month period from June to December 2023 that were collocated with Teledyne T640 FEM Monitors in two areas in Metro Manila: Pasig City and Manila. This study, using the same data, expands on this work by evaluating low-cost sensor calibration techniques using five machine learning algorithms - linear regression, support vector regression, random forest regression, XGBoost and gaussian model regression - and show significant improvement in the accuracy of low-cost sensor measurements. Performance was evaluated using metrics such as r-squared (R2), mean absolute error (MAE), and root mean squared error (RMSE). In the Pasig dataset, all ML models show a substantial increase in R² compared to the uncalibrated results. The uncalibrated data had an R² of 0.55284, while the ML models ranged from 0.77219 to 0.86749, representing an average improvement of 0.280562 and a percentage improvement of at least 39.68%. In the Manila dataset, the uncalibrated data had an R² of 0.61504 while having a very high RMSE of 23.02525. All ML models were able to bring the error down, ranging from 12.07652 to 19.40256 although linear regression did not improve with an R² of 0.58109 while Random Forest had shown significant improvement of 0.73457. These results suggest that even with the application of machine learning techniques, the inherent variability in sensor readings underscores the importance of sensor reliability for achieving accurate air quality assessments. Overall results also demonstrate that no single algorithm consistently outperformed others. Model performance was significantly influenced by factors such as data quality and the careful tuning of hyperparameters. The simplicity of linear regression and the flexibility of random forest may offer distinct advantages depending on specific application requirements. This study highlights the potential of machine learning for enhancing the accuracy and reliability of low-cost air quality sensor networks, paving the way for a more effective and affordable air quality monitoring solution for a country with limited air quality monitoring infrastructure like the Philippines.
   
• Sources and Elemental Composition of Fine Particulate Matter in Lancaster County and Implications for Public Health*
• Presented by: Ali Nawar, Franklin & Marshall College
  
  Air quality is a critical global concern, profoundly affecting human health and quality of life. High concentrations of fine particulate matter (PM2.5 and PM10) have a detrimental effect on the respiratory system and public health. Lancaster County has the 8th poorest air quality index (AQI) in the nation, primarily due to elevated levels of PM2.5 produced by various sources, possibly from coal-fired power plants, unregulated waste incineration, and automobile and diesel emissions, among others. Additionally, Lancaster County has a population of 552,597, with 13% affected by asthma, placing them at high risk for developing Chronic Obstructive Pulmonary Disease (COPD) and asthma-COPD overlap. This study assesses air quality by focusing on the composition of particulate pollution sources in Lancaster County, PA. Additionally, we are investigating Lancaster’s unregulated waste incineration through satellite monitoring of the area. These burning activities have been classified into three groups—burn pits, burn tanks, and burn barrels— based on size and distribution, and their spatial distribution has been documented via ArcGIS.
   
• Assessing the Health Impact of Indoor Air Purification Using Apple Watch Health Data: A Double Blind, Randomized, Crossover Study in the UAE*
• Presented by: Vince Nguyen, New York University Abu Dhabi
  
  Fine particulate matter (PM2.5) poses significant health risks. Indoor air purification has been shown to be effective at reducing PM2.5. However, most existing studies assess air purification’s health effects using specialized medical devices that are often expensive and impractical to scale. This study investigates short-term physiological responses to indoor air purification in real-world settings using the Apple Watch as a scalable, non-invasive, easy-to-implement, and high-frequency monitoring approach. Notably, it is the first study of its kind in the UAE, where ambient PM2.5 concentrations often exceed WHO guidelines. This study aims to assess whether the use of indoor HEPA air purifiers is associated with changes in cardiovascular, respiratory, and sleep health metrics obtained from existing Apple Watch wearers.
   
• Efficiency of low-cost sensors in monitoring the urban air quality impact of forest fires in coastal region of India*
• Presented by: Bharati Paul, Tezpur University
  
  Forest fires emissions significantly exacerbate air pollution by augmenting the atmospheric levels of primary and secondary pollutants, that are detrimental to human health and ecosystems, and degrade visibility. The air quality impacts occur through the emission of primary pollutants (e.g., PM2.5) and the production of secondary pollutants (e.g., O3). In this framework, this study provides a synthesis of the atmospheric fate of fire emissions and its impact on urban air quality in Andhra Pradesh, a coastal state in India. The study employs an integrative methodology to monitor the air quality during the active fire events in May 2024, incorporating ground measurements of Ozone (O3) and PM2.5 from low-cost sensors, fire and O3 data from MODIS, VIIRS and Tropomi; PM2.5 from MERRA-2 respectively. In addition, the study evaluates the efficacy of low-cost sensors by correlation analysis and error metrics to track urban air quality, particularly during active and post-fire events. The diurnal concentrations and spatial distribution of PM₂.₅ and O₃ over 15 locations (where low-cost sensors are installed) were analysed for the active-fire days and post-fire period. A comparative analysis was also executed between the low-cost sensors measured PM₂.₅ and O₃ concentrations and MERRA-2 PM₂.₅ and Tropomi O₃ datasets for the period.
   
• Introduction to Humidity-Resilient, Low-Cost Sensor Solution for PM Monitoring*
• Presented by: David Riallant, Groupe Tera
  
  The rapid urbanization and industrialization of Southeast Asia have raised growing concerns about air quality, particularly regarding fine particulate matter (PM1, PM2.5, and PM10).While low-cost sensor (LCS) technologies offer a scalable solution for widespread air quality monitoring, their deployment in this region presents unique challenges—primarily due to high humidity levels that compromise data reliability, leading to inaccuracies, sensor drift, and premature aging. In this talk, we will introduce a next-generation optical particulate matter sensor specifically designed to overcome these environmental constraints. Unlike conventional low-cost sensors, which suffer from hygroscopic particle growth and degradation over time, the NextPM sensor integrates a patented humidity regulation system. This innovation ensures stable and accurate measurements even in high relative humidity conditions, preventing data drift and sensor clogging. Combined with its dual detection angle (measuring light scattering at two angles, 45° and 90° ) NextPM also improves measurement precision to offer an overall accurate, reliable and yet affordable dust monitoring solution. We will present a comparative analysis of sensor performance—with and without humidity regulation—against reference instruments, along with statistical insights demonstrating the effectiveness of this technology in humid environments. NextPM has been granted Class 1 Certification in Korea, a region with similarly high humidity conditions. With an accuracy of 80% or higher compared to official reference instruments, the sensor is validated for use in both regulatory and scientific applications, offering a reliable solution for improved air quality monitoring in challenging climates.
   
• The Air Aware Oakland Workforce Development Program: Training the Next Generation of Air Quality Technicians*
• Presented by: Jeff Sanchez, Sequoia Foundation
  
  The Air Aware Oakland Workforce Development Program provided low-income Oakland, California youth with hands-on tools and experiences to site and install air sensors. Over five weeks, interns participated in a comprehensive training program that combined technical education, community engagement, and professional development. Designed to integrate equity and environmental justice principles, the program’s day-to-day activities empowered interns to actively contribute to air quality monitoring efforts in two high-priority/AB 617 communities (East and West Oakland) while building skills for future career paths. This presentation will provide an in-depth look at the program’s structure, highlight the role of external partnerships and technical training in advancing equity, discuss how such initiatives can be scaled or adapted for other communities, and hear about the challenges we faced. Attendees will gain insights into effective strategies for fostering community-driven solutions to environmental health challenges while building the capacity of underserved populations.
   
• Evaluating the impact of real-time, public air quality informational displays: unpacking the AWAIR Project*
• Presented by: Kayla Schulte, Imperial College London
  
  Air pollution remains a critical public health challenge, especially in urban areas, where residents face high levels of harmful pollutants. Despite this, many communities lack access to real-time information that could help mitigate air pollution exposures. The AWAIR project addresses this gap by providing co-created, accessible air quality information in real-time via public displays in three neighborhoods across London, UK. AWAIR aims to support and work with communities by offering air pollution information that can reduce personal exposure and stimulate community-driven efforts to lower emissions. The project also explores how real-time information can inform broader public health strategies and guide future interventions.The methodological approach for this project involved identifying three London neighborhoods or ‘wards’ for the pilot display deployments using social and environmental datasets, while working alongside local community groups. A baseline survey was conducted to evaluate residents’ existing knowledge, attitudes, and behaviours related to air pollution. The survey also examines how sociodemographic factors such as age, gender, income, and education correlate with awareness levels. This was followed up by co-design workshops with residents to ensure the displays were relevant, accessible, and culturally appropriate. The workshops led to the creation of a new air quality communication scale and messaging that incorporates both the World Health Organization (WHO) guidelines and the UK’s Daily Air Quality Index (DAQI).A second survey was then conducted to evaluate changes in knowledge and behaviour. Logistic regression analysis is used to identify key social and environmental factors associated with display interaction and uptake of exposure reducing behavioural recommendations delivered through the display. Comparative analysis with national datasets supports our evaluation of whether local engagement patterns with air pollution information align with national trends.Overall, this study provides valuable insights into the efficacy of community-guided air quality displays in raising public awareness and supporting exposure reducing behaviour. By comparing local and national trends, the project aims to inform future interventions, equipping policymakers and public health officials with the tools to better tailor strategies to different demographic groups. Ultimately, this research contributes to the broader goal of reducing public exposure to harmful air pollutants and improving health outcomes at the local level.
   
• A Model for Reconstructing Personal PM2.5 Exposure Using Low-Cost Sensors and CAAQMS*
• Presented by: Kirtika Sharma, Indian Institute for Technology Delhi
  
  Accurately quantifying personal exposure to PM2.5 is crucial for understanding its health impacts. In India, the Continuous Ambient Air Quality Monitoring Systems (CAAQMS) provide precise measurements, but their sparse spatial distribution and high deployment and maintenance costs limit their effectiveness for exposure assessments at the individual level. Low-cost sensors (LCS) offer greater spatial coverage but are often affected by data quality issues, necessitating rigorous calibration. These challenges collectively hinder the development of high-resolution exposure estimates critical for advancing air pollution research and public health assessments.To address this gap, we developed a framework by integrating calibrated static LCS data with CAAQMS observations to generate high-resolution (1 km × 1 km) and 1-hour PM2.5 estimates. Our study focuses on Kolkata, India, during the winter season (December 1, 2023 - January 31, 2024), leveraging PM2.5 data from seven CAAQMS stations and 22 static LCS stations. LCS calibration was performed using meteorological variables—temperature and relative humidity—sourced from the nearest CAAQMS stations.We employed multiple machine learning models to integrate and harmonize the datasets, effectively capturing complex non-linear relationships and enhancing predictive accuracy. The results demonstrate that integrating LCS data significantly enhances model performance by reducing RMSE across all methods. Random Forest showed the greatest improvement, with a 23.63% reduction in RMSE while maintaining a high R² of 0.90. XGBoost also benefited from LCS integration, achieving a 7.46% reduction in RMSE, though its R² decreased from 0.91 to 0.85. These findings underscore the effectiveness of a hybrid framework that integrates LCS with CAAQMS data for generating gridded PM2.5 data at very high spatial and temporal scales. This data can further be utilized to reconstruct personal PM2.5 exposure by time-weighted average based on their geolocations.
   
• PM2.5@Asia: an education-oriented project based on a research-grade regional PM2.5 monitoring network*
• Presented by: Sheng-Hsiang Wang, National Central University, Taiwan
  
  Air pollution poses significant challenges to public health, environmental sustainability, and climate change, particularly in Southeast Asia. To address these issues, we have established a monitoring network, PM2.5@Asia, which focuses on PM2.5 and CO2 monitoring, leveraging advanced low-cost sensor technologies. A key component of this initiative is the Aerobox, a cost-effective and research-grade sensor system designed for air quality monitoring. In 2019, five Aerobox units were deployed in Chiang Mai, Thailand, where their performance and accuracy were evaluated against standard air quality monitoring stations. The results demonstrated strong agreement, validating Aerobox as a reliable tool for air quality assessment. Supported by a University Social Responsibility (USR) project, PM2.5@Asia has since grown into a regional network, fostering collaboration among universities to promote education, outreach, and research on air pollution and its impacts. Currently, 10 universities across Southeast Asia are participating in PM2.5@Asia, working together to raise awareness and develop solutions for air quality challenges. This presentation will provide an overview of the network, highlight recent progress, and explore the research potential of Aerobox and PM2.5@Asia in advancing air pollution education, transboundary pollution studies, and climate change mitigation efforts.
   
• Indoor and Outdoor Aerosol Sources and Their Effects on Air Quality in Changchun, China*
• Presented by: Yuliia Yukhymchuk, Jilin University
  
  The air quality in Northeast China is impacted by a range of aerosol sources, including both outdoor and indoor pollution. Outdoor sources comprise anthropogenic emissions from industrial activities, transportation, and natural aerosol events such as mineral dust transfer. Indoor pollution is primarily attributed to domestic cooking and specific ventilation issues in residential buildings. During the Chinese New Year holidays, a high fireworks volume significantly deteriorates Changchun's air quality. Additionally, extensive cooking activities during the holiday season lead to substantial indoor particle emissions, often causing PM2.5 concentrations indoors to exceed outdoor levels. In the spring, air mass movements from the Gobi and Taklamakan deserts contribute to further air quality degradation outdoors and indoors. This study investigates the effects of holidays, such as the Chinese New Year, and natural aerosol sources, such as mineral dust transfer from deserts, on air quality in Changchun. The research focuses on concentrations of PM2.5 and PM10, as well as the properties of aerosols. Data was collected using low-cost air quality sensors (AirVisual Pro, part of the AirVisual Network) and information from the CIMEL sun lunar sky photometer recently installed at the Changchun_JLU AERONET station. The findings highlight the complex interactions between anthropogenic and natural factors that contribute to the region's air quality variability.
   
• Assessment of Transboundary Air Pollution in Bangladesh: Influence of Boundary Layer Height and Local Winds Using Low-cost Sensor Network*
• Presented by: Shaid Uz Zaman, Bangladesh University of Engineering and Technology
  
  Low-cost sensor (LCS) networks offer spatially resolved information on PM2.5 variations in urban areas, addressing the significant challenge of air pollution, which has severe environmental and societal impacts, especially in developing countries like Bangladesh. We have assessed the impact of transboundary air pollution using a network of LCS over Panchagarh, Rajshahi, Dhaka, and Bhola from April 2022 to September 2023. A generalized additive model (GAM) was developed to analyze the hourly data for both dry and wet seasons. Significantly high concentrations were observed for all the sites surpassing the guidelines set by both World Health Organization (WHO) and Department of Environment (DoE), Bangladesh. The highest overall PM2.5 concentrations were observed in Panchagarh (69.8 ± 56.8), followed by Rajshahi (54.0 ± 33.7), Dhaka (53.5 ± 38.8), and Bhola (44.5 ± 37.3). The GAM results showed that the lowest contribution of boundary layer height is at rural Bhola and Panchagarh, and highest at Dhaka. The contribution of long-range transport was found uniform at all the sites. The Trajectory Cluster Concentration Impact (TCCI) showed that the Indo-Gangetic Plain (IGP) is responsible for the enhancement of 40 μgm-3 at all the sites. However, wind transported from the Bay of Bengal associates PM2.5 reduction of 20 to 40 μgm-3. Impacts of local winds on the PM2.5 concentrations in the GAM simulations suggested that winds from the northwest are associated with higher PM2.5. This study underscores the importance of a dense LCS network to effectively monitor and address the severe effects of transboundary air pollution in Bangladesh.
   
• Development of Bamboo-Derived Activated Carbons for Volatile Organic Compound (VOC) Sensing: An Electrochemical Impedance Spectroscopy Study
• Presented by: Jon Bell, Zurich University of Applied Sciences
  
  Bamboo-derived activated carbon (BDAC) is a highly porous material derived from the pyrolysis and chemical activation of bamboo. Bamboo is a CO₂ net-zero material due to its rapid growth rate and CO₂ offsetting potential and therefore forms a major part of Thailand’s Bio-Circular Green Economy (BCG) Model. Activated carbons are excellent at adsorbing a wide range of volatile organic species with different physicochemical characteristics due to their wide pore size distribution and hydrophobic and hydrophilic adsorption sites. Planar hydrophilic VOC species, such as aromatic compounds, are usually adsorbed in hydrophobic microporosity, while polar species, such as water, methanol, and ethanol, are adsorbed on oxygen functional groups at the edge of graphitic layers. The adsorption of polar and non-polar VOCs can significantly alter the resistive and capacitive characteristics of the carbon, which may also have an A.C. frequency dependence due to differences in diffusion timescales for species adsorbed in different pore sizes. In this study, bamboo was pyrolysed at both 500 °C and 800 °C using indirect and direct heating methods, respectively, to produce two biochars, which were then chemically activated in KOH and heat-treated in air at 700 °C to develop the microporous structure for enhanced VOC adsorption and to produce the BDAC material. These carbons were then milled and sieved to produce particles between 125 and 250 microns. The porous structure was characterised using N₂ adsorption at 77 K and CO₂ adsorption at 273 K to investigate the total pore and micropore volumes, respectively. Initially, vapour adsorption properties of the powder were studied for benzene and water vapour over the temperature range 25 °C to 45 °C using gravimetric dynamic vapour sorption (DVS) instrumentation. Following this, the powder was pressed into a pellet, and silver electrodes were formed on the faces for impedance spectroscopy vapour sensing studies. The Ag-BDAC-Ag pellet was placed inside the DVS system to control the vapour composition and temperature. The vapour adsorption isotherms and A.C. electrical properties were measured simultaneously, as a function of temperature and benzene and water vapour partial pressure, to investigate the DBAC vapour detection properties of non-polar and polar molecules. Therefore, with this experimental setup, the electrochemical impedance response can be directly related to the vapour adsorption isotherm and adsorption kinetics. This work has produced a sustainable carbon material for use in vapour sensing and air purification systems. Additionally, the vapour detection properties of the BDAC could be used for breakthrough and saturation detection in BDAC packed bed air purification systems. Therefore, this work provides a solid basis to develop the next generation of smart air purification systems.
   
• Assessment the impact of abating dispersed sources on local air quality using low-cost sensors
• Presented by: Sagnik Dey, Indian Institute of Technology Delhi
  
  Air pollution is the most significant environmental health risk in India. To combat air pollution, India has launched the National Clean Air Program. However, due to an inadequate ground monitoring network of reference-grade monitors, detecting changes in local air quality after specific measures are implemented at hotspots poses challenges.  In our study, we selected three regions—Rohini, Jahangirpuri, and Karol Bagh in New Delhi—and deployed 10 well-calibrated low-cost sensors in each region within a 1-km radius. Over our five-month deployment period, local authorities resolved 66 long-term issues in these areas, such as paving unpaved roads and filling potholes, clearing garbage and debris, cleaning overflowing trash bins, greening barren lands, and repairing broken footpaths.  By applying a difference-of-differences approach and removing the seasonal influence of the shallowing boundary layer on PM2.5 concentrations, we found that PM2.5 levels decreased by 15.3% in Karol Bagh, 26.6% in Jahangirpuri, and 15.7% in Rohini due to the local interventions. The magnitude of these impacts depends on pre-intervention PM2.5 levels, the duration and scale of the interventions, and the presence of other local sources. Our study demonstrates the utility of hyperlocal monitoring through sensor technology in evaluating the impact of dispersed source management programs, enabling local administration to make informed decisions. Cities with sparse ground-based monitoring networks can adopt this approach to evaluate their local air quality management plans.
   
• Public awareness and exposure assessment of street fruits vendors to traffic related atmospheric pollution
• Presented by: Anil Namdeo, Northumbria University
  
  Africa is home to most developing nations and home to some of the cities growing at a fast pace. The accelerated demographic growth and the industrialization are considered as the main key drivers of the observed urbanization. The informal economic sector plays a major role in the economy of most places where street vendors are found on roadsides in every city, selling various types of goods. Those dealers spend the entire day on nearby roads and are exposed to pollution from motor vehicles. Nevertheless, very few studies have focused on this category of workers to examine the level of awareness and information related to the risks they are exposed to. In this study, fruits dealers were targeted. In the city of Thiès, in the Western region of Senegal, a walkthrough survey was conducted to map the locations of these vendors, evaluate their awareness and assess their exposure to traffic related pollution. A portable air quality monitor, the Particles Plus 8301-AQM2 Series was used monitor particles concentrations. Within a stretch of 10 km, more than 50 fruits vendors were recorded, and the survey was carried with 35 of them. Exposure to particles was evaluated for 10 vendors at selected locations. 86% of the vendors were aged 30-50, and 65% of the vendors were female. 94% of the vendors didn’t know what air pollution is, but more than 80% recognized dust and exhaust gases as harmful substances. More than 80% never heard any information related to air pollution, nor knew there was an air quality management centre in the country. For all locations PM2.5 concentrations exceeded the threshold value of 15 μg/m3, from two up to five-fold. PM10 concentrations exceeded 45 μg/m3 and the averages recorded were in the range 250-700 μg/m3. The findings point out the need to raise awareness of air pollution amongst streets dealers, inform and help them monitor their health and safety.
   
• Ultrafine Particles from 3D printing in an Equipment Assembly Facility
• Presented by: Anil Namdeo, Northumbria University
   
  This study reports on ultrafine particle emissions from 3D printing facility in an equipment assembly facility in USA. 3D printing, while innovative and versatile, does come with some environmental concerns, particularly regarding air pollution. The process of 3D printing involves melting plastic filaments to create objects layer by layer. This heating process releases various emissions, including volatile organic compounds (VOCs) and ultrafine particles. Ultrafine particles are especially concerning because they are small enough to penetrate deep into the respiratory system and can pose health risks when inhaled. Studies have shown that materials commonly used in 3D printing, such as acrylonitrile butadiene styrene (ABS) and polylactic acid (PLA), can emit these particles. ABS, in particular, tends to release more particles compared to PLA. Two optical particle counters (make Particles Plus) and two water-based condensation particle counters (make Particles Plus) were place in a room where a 3D printer is installed to produce various parts required in an equipment assembly unit. The study was conducted for a total of 132 h in a working week. Within this period 3D printer was used for a duration of 36 hours. OPC recorded particles size distributions and counts in the range of 300 nm to 10 μm. Water-based condensation particle counter provided total particle number counts above 50 nm. The results of this case study have provided a better understanding of the particle size and number distributions from a 3D printing facility. This will add to the knowledge in this field and help in developing solutions to minimize emissions of ultrafine particles and ways to reduce the exposure of the workers using 3D printers. To mitigate the exposure risks, it's essential to use 3D printers in well-ventilated areas and consider printers with built-in air filters. Additionally, using safer materials and adopting best practices for handling and operating 3D printers can help reduce emissions.
   
• Application of optical particle counters to understand spatial variations of PM2.5 and PM10 in outdoor and indoor environments in Mexico City
• Presented by: Elizabeth Vega, UNAM Mexico
  
  Mexicans spend nearly 90% of their time indoors, potentially having significant exposure to air pollutants including particles (PM). Levels of PM2.5 and PM10 are frequently above the Mexican air quality standards and the WHO guideline limits. Two optical particle sensors were simultaneously placed, in rotation, in 38 homes for seven days for measuring indoors and outdoors particle size and number concentrations. PM were measured in six size ranges from 300 nm to 10 mm. This provided a comparison between indoor and outdoor PM levels and size distributions. This presentation focuses on PM2.5 and PM10. The surveyed 38 homes covered all municipalities of Mexico City and provided a good understanding of spatial variations in indoor and outdoor PM2.5 and PM10 concentrations and size distributions. Each participant completed a diary of relevant conditions, with information such as ventilation level, cooking activities, fires for heating, lighting candles, or burning incense. Participants also completed an online questionnaire to capture broader details, such as the number of occupants in the home, the type of building construction, and proximity to industry and heavy traffic. The PM2.5/PM10 ratio was, on average, 0.3, both indoors and outdoors. 21% of the analyzed homes were observed to have a higher maximum PM2.5 concentration outdoors than indoors (with a maximum of 3.6 times the indoor concentration). In 18% homes, these concentrations were similar (proportions of 0.8 to 1.2), and in 60% of cases, the maximum PM2.5 concentration was higher indoors (up to 9 times the maximum outdoor concentration). Higher indoor PM2.5 concentrations were found to be primarily related to cooking and, to a lesser extent, cleaning activities.
   
• Bias Correction of Air Quality Forecasts using Machine Learning
• Presented by: Mohammad Rafiuddin, Council on Energy, Environment & Water
  
  Air quality forecasting plays a critical role in preemptive air pollution management. These forecasts are generally generated using Chemical Transport Models (CTMs) that not only give information around pollutant concentration but also the contribution of sources both spatially and temporally.  While CTMs provide valuable insights, they are prone to errors rising due to various reasons, like uncertainties in emission inventories and inaccurate weather forecasts. As a result, their predictions can deviate significantly from observed air pollution levels, especially during extreme pollution events. Recently, machine learning models have shown significant potential in correcting these biases as they are able to capture the non-linear interactions between emissions and meteorological factors. We apply three ML models—Random Forest, Gradient Boosting Decision Trees (GBDT), and XGBoost—to correct biases for daily PM2.5 forecasts. Our approach is tested on a high-resolution (1×1 km) CTM based on WRF-CAMx for the Delhi-NCR region, as well as global forecasting systems such as SILAM and GEOS-CF for Mumbai. These models take raw CTM predictions of PM2.5 with meteorological variables like temperature, rainfall, and wind speed as inputs. The corrected forecasts are then validated against ground-based air quality monitoring station data, and finally the best model is chosen.Results show a significant improvement in forecast accuracy, with GBDT outperforming the other models. Bias correction reduced the Mean Absolute Percentage Error (MAPE) from 46% to 24% for WRF-CAMx, 114% to 26% for SILAM, and 88% to 35% for GEOS-CF. However, the model tends to underestimate extreme pollution events. Future work will address this limitation by implementing techniques such as SMOTE for regression, relevance functions, and customized loss functions that penalize large errors more effectively. The proposed ML-based bias correction method is computationally inexpensive, easy to implement, and scalable to different cities using existing global air quality forecasts.
   
• Weak enforcement of air pollution regulations: A case for expanding air quality monitoring
• Presented by: Hrishikesh Chandra Gautam, Air Quality Life Index at UChicago
  
  Air pollution reduces global life expectancy by an average of two years. Among the key determinants of pollution levels in a country are air quality standards and the ability to enforce them.In this study, we apply the Air Quality Life Index (AQLI) to particulate pollution data from 252 countries to estimate potential life expectancy gains if pollution levels met WHO guidelines. We also assess the global landscape of air quality standards and compare pollution impacts between countries with and without such standards.Our findings indicate that people in countries with stricter air quality standards (less than WHO interim target-3 of 15 µg/m3) live 1.6 years longer, on average, than those in countries with weaker or no standards. Despite the importance of these standards, only 94 of 252 countries have adopted them, and just 57 of these 94 countries successfully meet their own standards.Air quality monitoring is also closely linked to the presence and enforcement of standards. Countries with stricter air quality regulations have 14.3 monitors per 100,000 people, compared to 2.6 per 100,000 in those with weaker regulations. Notably, among the 158 countries without air quality standards, 105 have no evidence of air quality monitoring. Moreover, among the 37 countries that fail to meet their own standards, 18 have fewer than 5 (11 have fewer than 1) monitors per 100,000 people, compared to those that successfully meet their standards.These results highlight the critical role of air quality data in helping countries establish and enforce stricter pollution standards, ultimately improving public health outcomes.
   
• Assessing the Impact of Biomass Burning in Haryana on Delhi’s Air Quality Using a low-cost Sensors Network
• Presented by: Amarendra Singh, Indian Institute of Technology Delhi
  
  Air pollution poses a major public health risk, particularly pronounced in a densely populated metropolitan cities like Delhi where vehicular emissions elevate PM2.5 levels. In developing countries, regulatory monitoring stations provide accurate data but are expensive and sparse. Low-cost sensors (LCS) offer a cost-effective solution, capturing fine-scale spatial and temporal PM2.5 variations. This study evaluates pollution control measures using a strategically deployed LCS network in Haryana and Delhi with robust calibration process by collocating and calibrating the sensor with Beta Attenuation Mass (BAM) located at the IIT, Delhi testbed. The objective is to understand the changes in PM2.5 levels in upwind, downwind and within the Delhi NCT during biomass burning (BB) episodes. To investigate this, we deployed 33 LCS across 21 locations along a 140 km transect (315°-322° wind direction) toward Delhi. Sensors were spaced 10 km apart, except in Delhi, where two sensors were deployed within 500 m (upwind and downwind) to minimize data loss. PM2.5 concentrations were measured from September to November 2024 to analyse the impact of BB in Haryana on Delhi’s air quality. Results show that upwind locations recorded higher PM2.5 levels in October (158 μg/m³) and November (333 μg/m³) compared to Delhi (147 μg/m³ and 272 μg/m³). Downwind concentrations matched Delhi’s, at 152 μg/m³ and 272 μg/m³. During the high pollution episode (13-23 November), upwind, Delhi, and downwind PM2.5 concentrations were 407 μg/m³, 362 μg/m³, and 334 μg/m³, respectively. During October, the highest peak in pollution levels for upwind and Delhi was observed at 6:00 AM, while for downwind areas, it occurred at 7:00 AM. In November, the peak pollution for both upwind and downwind regions was recorded at 7:00 AM, whereas in Delhi, it was observed earlier at 3:00 AM. During the high pollution loading period (13-23 November), the peak timings shifted significantly, with upwind areas peaking at 2:00 AM, Delhi at 5:00 AM, and downwind areas at 4:00 AM. These variations highlight the influence of BB on pollution dynamics in Delhi.
   
• Characterizing Seasonal and Diurnal PM2.5 Patterns in Surat using LCS: Implications for Public Health and Urban Policy
• Presented by: Kiran Suryawanshi, SVNIT Surat
  
  Rapid urbanization and industrialization have intensified air pollution in cities like Surat, India, leading to significant public health and environmental challenges. This study used data from a network of 20 low-cost sensors (LCS) across Surat to monitor PM2.5 concentrations over one year. Through spatiotemporal analysis, we identified pollution hotspots and assessed exposure levels across various city zones. The Multiple Path Particle Dosimetry (MPPD) model was utilized to analyze PM2.5 mass deposition across different spatiotemporal dimensions. Findings revealed distinct seasonal and diurnal patterns in PM2.5 concentrations, with industrial and high-traffic zones experiencing critical deposition levels. Seasonal spikes were observed during winter, attributed to reduced atmospheric dispersion, while higher concentrations during peak traffic hours underscored the impact of vehicular emissions. These insights informed targeted mitigation strategies, including optimizing industrial emissions control, enhancing urban green infrastructure, and implementing traffic management measures. The study underscores the potential of low-cost air sensors as effective tools for real-time air quality monitoring and exposure assessment in rapidly urbanizing cities. By integrating sensor-based findings into public health policies, we can develop sustainable urban planning strategies to mitigate air pollution and protect public health.
   
• Characterization of indoor and outdoor PM2.5 levels in residential areas of Ho Chi Minh City using low-cost sensors
• Presented by: Thi Hien To, University of Science, Vietnam
  
  This research provides information on indoor and outdoor PM2.5 concentration in the urban atmosphere of Ho Chi Minh City (HCMC), Vietnam. PM2.5 concentration data were collected daily at five locations, including three residential sites (Q6, Q9, QNB) and two roadside sites (Q12, QBT) within Ho Chi Minh City from 5/17/2021 to 5/23/2021. Data were collected every 15 seconds using PM2.5 low-cost sensors. The sensor was positioned 1.5 m above the ground. The daily average indoor PM2.5 concentration at the 5 locations were respectively 19.76 ± 14.28, 21.46 ± 12.75, 20.18 ± 11.64 μg/m3 at residential sites and 13.89 ± 12.62, 19.71 ± 11.20 μg/m3 at the two roadside sites. The daily average outdoor PM2.5 concentrations are respectively 19.28 ± 12.84, 18.11 ± 11.47, 25.72 ± 18.69, 13.39 ± 9.53, 19.44 ± 15.39 μg/m3. Indoor PM2.5 concentrations at the five locations peaked on Wednesday and decreased on the weekend, similar to outdoor PM2.5 concentrations during that week. Indoor PM2.5 concentrations at all five locations met Vietnam's indoor air quality guideline (TCVN 13521:2022). Similarly, outdoor PM₂.₅ concentrations at all locations complied with Vietnam's National Ambient Air Quality Standards (NAAQS) (QCVN 05:2023). The indoor to outdoor (I/O) ratios of PM2.5 were 1.1, 1.3, 0.9 at three residential sites and 1.6 at the roadsides. The I/O ratios greater than 1.0 indicated that indoor sources such as incense burning, cooking, and tobacco smoke contributed to the elevated indoor PM2.5 concentrations.
   
• Risk perception and behavior changes among rural tribal households participating in novel air quality monitoring with varied report-back methodologies
• Presented by: Scott Collingwood, University of Utah
  
  members were recruited to voluntarily participate in in-home air quality monitoring (aerosol). Households were randomly assigned to use a small, novel particulate monitor giving them real-time feedback of PM2.5 or one that provided no aerosol exposure estimate information. Environmental Health Perception and Behavior (EHPB) questionnaires were distributed to participants before and after monitoring equipment was placed in participants’ homes. No significant changes in behavior, perception, or concern occurred among the participants. Households who had a feedback monitor in their home were significantly less likely to complete the follow up questionnaire. Due to the small sample size, changes in concern, thoughts, or behaviors were difficult to detect. Suggestions for future work are discussed, including; overcoming challenges of in-home research activities in the COVID-19 era, bilateral participant communication for study operations and questionnaire/survey completion, as well as improving the return of research results for tribal participants.
   
• Openair quality monitoring and transparent data sharing in Malawi: The Malawi Initiative for Clean Air Solutions (MIfCAS)
• Presented by: Collins Gameli Hodoli, Clean Air One Atmosphere
  
  Due to the significant impact of poor air quality on human health, there has been a global drive for country-specific air quality monitoring infrastructures. This includes efforts to transparently share generated data for domain-specific research targeted at protecting public health and environmental resources. The broad majority of countries in Africa, including Malawi, have limited measurement capacity in part due to logistical and human capacity issues associated with operating air quality monitoring systems and their associated data management infrastructures. The Malawi Initiative for Clean Air Solutions (MIfCAS) is a collaborative project between local government agencies (Malawi Standards Bureau, Ministry of Natural Resources and Climate Change, and Environmental Protection Authority), local research institutions (Mzuzu University, and Malawi University of Science and Technology, Malawi) and international partners (Imperial College, London). Funded by the Energy Policy Institute of Chicago’s 2025 Air Quality Fund, this work aims to support national policy formulation by bridging data gaps and present a model for country-specific strategies for air pollution management and control specifically for environments previously not monitored. The aims of this project are to generate (i). an observational model useable in similar environments in other parts of Africa, (ii). a guide on establishing air quality monitoring stations in the region and (iii). documentation on transparent infrastructures for open data/knowledge sharing. MIfCAS will build an air quality infrastructure in Malawi that is locally owned and sustained where the generated data will be used to initiate/support a national discourse for clean air solutions in Malawi.. The generated data from all these stations will be shared via the OpenAQ air quality data repository. In the long-term, MIfCAS will support development of air pollution policy in Malawi including ambient air quality standards. This will help reduce the current estimated annual 15,900 premature deaths in Malawi associated with exposure to air pollution using domain specific data. A key part of this project will be better characterization of local pollutant distribution and sources, supporting development and implementation of district-specific air quality management plans (for the 28 local districts in Malawi) and building in-country capacity for air quality monitoring and management.
   
• Performance of Commercially Available CO₂ Gas Monitor for Air Quality Monitoring: A Comparative Study*
• Presented by: Jaafar Nur Akasyah, International Islamic University Malaysia
  
  This study presents a comparative performance of four commercially available non-dispersive infrared (NDIR) CO₂ gas monitors, named gas monitor (GM) 1, 2, 3, and 4, each utilizing distinct sampling methods: active (with pumps) and passive (natural diffusion) sampling. A 148.19 L test chamber was developed in accordance with ISO 26142:2010 to ensure controlled and reliable conditions. Gas monitors were evaluated for six performance parameters: accuracy, stability, selectivity, repeatability, response time, and recovery time. Results indicate significant variation among gas monitors, largely attributable to their sampling strategies. GM 3 showed the highest accuracy (±1.2% deviation from 808 ppm), GM 4 demonstrated superior stability (standard deviation of ±5 ppm over 1 hour), while GM 1 and GM 2 recorded the fastest response and recovery times (0.9 min and 0.1 min), respectively. The findings reveal notable differences in gas monitor behaviour linked to their detection principles and sampling strategies. This study provides valuable insights into selecting the most suitable NDIR CO₂ gas monitors for targeted air quality monitoring applications.
   
• Quantification of residential particulate pollution using low-cost sensors and the perception of air quality in Ibadan, Nigeria*
• Presented by: Jim McQuaid, University of Leeds