LocationsSystems for Clean Air: Connecting Data, Innovation, and Institutional Change
June 10, 2026
Researchers, policymakers, funders and practitioners at a roundtable co-hosted by the UNDP Global Centre Singapore and the Centre for Climate Change and Environmental Health (CCEH) at NTU.
Air pollution remains one of the world’s most pressing environmental health threats, yet paradoxically, the countries most burdened by it have the least capacity to measure it. With 36% of governments not conducting air quality monitoring and low and middle-income countries (LMICs) bearing the greatest health burden, a critical gap persists between the scale of the problem and the ability to track it. While advances in low-cost sensor technology and AI have made hyperlocal, dense monitoring financially and technically feasible, a deeper challenge remains: translating monitoring data into scaled action across the multiple sectors and institutions that determine air quality. This requires having appropriate governance structures and cross-sector coordination mechanisms in place that provides the necessary evidence needed to drive policy.
To examine this challenge and discuss potential solutions, researchers, policymakers, funders and practitioners convened at a closed-door roundtable co-hosted by the UNDP Global Centre for Technology, Innovation and Sustainable Development and the Centre for Climate Change and Environmental Health (CCEH) at NTU. The discussion, titled Systems for Clean Air: Connecting Data, Innovation, and Institutional Change, focused on what it takes to bridge the gap between monitoring data and policy, across three interconnected dimensions:
- The evidence base linking air pollution and climate to health outcomes;
- The monitoring and innovation infrastructure needed to generate credible, decision-ready data; and,
- The governance structures and cross-ministry coordination mechanisms that determine whether data translates into action
The CCEH at NTU has been conducting research to expand the evidence base. Professor Steve Yim, Director of the CCEH, presented research on how AI can be applied to sensor data to generate high-resolution maps of pollution sources, forecast transboundary haze events two to three days in advance, and to produce early health warnings broken down by disease type and location.
“With detailed understanding of the characteristics of air quality and weather, we can develop models to predict, forecast, and monitor episodes, and to also explain what happened during an episode and when it would disappear.”Professor Steve Yim, Director of Centre for Climate Change and Environmental Health (CCEH) and Professor of Asian School of the Environment and Lee Kong Chian School of Medicine
While CCEH’s research illustrates how far science has advanced, the capacity to generate this kind of evidence remains unevenly distributed. Secondary and tertiary cities, rural areas affected by agricultural burning, and remote communities across LMICs rarely have a single monitor within reach. Where data is absent, policy cannot follow.
The Air Pollution Monitoring Gap
Air pollution is one of the biggest environmental risks in the world, yet its monitoring remains severely inadequate relative to the burden it is meant to address. As discussed above, there is a critical gap between the scale of air pollution in LMICs and the ability to track the issue. Although LMICs bear the greatest burden of air pollution, they are the least equipped to measure it. This gap is made more acute by the recent shutdown of U.S. embassy air quality monitors, which led to 44 countries being affected overnight.
Without data, policymakers can neither identify areas where pollution is worse, nor set evidence-based standards, design targeted interventions, and measure whether policies are working. This creates a negative self-reinforcing cycle: no data means no policy, no policy means no investment, and no investment means no data.
Low-Cost Sensors (LCS) and Their Potential
Within the right institutional context, LCS can serve as a potential solution to the problem. Previous articles in our blog series have documented how the economics of LCS have shifted dramatically, how calibration techniques and data fusion methods can substantially improve their accuracy, and how communities with no prior technical background have successfully deployed and managed sensor networks. Hyperlocal, dense air quality monitoring is financially and technically within reach of LMICs.
The EU’s 2024 Ambient Air Quality Directive represents a turning point in recognizing this shift. For the first time, LCS are recognized as a legitimate source of indicative measurement, supplementing reference-grade monitors. For LMICs, the directive sets a precedent that regional frameworks are likely to follow.
The Open-Source Air Quality Monitoring Toolkit
Recognizing both the potential and the limitations of LCS, UNDP has developed practical responses addressing two of the most persistent barriers: the lack of accessible network design guidance for communities with limited technical capacity, and the absence of institutional coordination mechanisms that connect monitoring data to policy action.
Co-developed by UNDP GC-TISD and AirGradient, this toolkit provides communities with no technical background step-by-step guidance to plan, set up, and manage an affordable monitoring network. A pilot in Hanoi saw over 70% of non-technical users operating devices independently. To make the Toolkit more intuitive and easier to use, an AI-powered conversational chatbot is under development that will allow users to interact with the toolkit in plain language, providing the most relevant answers to their specific queries and referencing sections of the toolkit for additional guidance.
The Open-Source Toolkit addresses a key technical barrier by democratising access to hyperlocal air quality data, driven by the growth, technological improvement, and increasing recognition of low-cost sensors. Yet the monitoring network alone is not enough. The institutional and coordination arrangements that translate data into better air quality outcomes are equally essential.
“As technologies like AI increase the accuracy and spatial detail of air quality information, the channels that convert data into action become more important, not less. Better data without the institutional arrangements to act on it does not change policy.”Carla Gomez, Sustainability and Climate Specialist, UNDP Global Centre Singapore
A Systems Approach to Clean Air: The Clean Air Systems Innovation Playbook
Advances in monitoring technology and regulatory shifts are broadening the possibilities. However, these opportunities coexist with a structural problem: air quality is determined by decisions made in a siloed manner, such as within the planning, finance, transportation, industry, energy, health, and agricultural ministries, with each ministry having their own mandates and goals. Without a horizontal framework that aligns these mandates with national-level goals, technology and data remain disempowered to drive policy.
UNDP’s Regional Bureau for Asia and the Pacific is working on a Clean Air Systems Innovation Playbook using a cross-ministerial portfolio approach to shift air quality from a fragmented environmental issue across ministries to a shared and coordinated government-wide outcome. It engages line ministries in a collaborative workshopping process to map where mandates overlap, identify where finances and field resources are duplicated, and create a concrete portfolio of joint efforts. The aim is to make clean air a cross-government priority rather than a single ministry’s duty, just as gender-responsive budgeting changed how governments prepare for gender parity.
“Air pollution is everyone’s problem, but no one’s KPI. The drivers sit across multiple ministries, budgets, and mandates. A systems approach becomes inevitable when no single institution controls the outcome.”Aafreen Siddiqui, Regional Engagement Lead, Digital, Public Sector Innovation and Partnerships, Regional Innovation and Digital Team, UNDP Regional Bureau of Asia Pacific (Bangkok Regional Hub)
The Institutional Challenge Ahead
There is no single lever that unlocks clean air. Better science, more accessible monitoring, and stronger governance are each necessary, but none is sufficient on its own. Progress on one dimension, without the others, reaches its limits quickly. What shifts outcomes at scale is a connected set of efforts across sectors and over long horizons, where each piece reinforces the others. This requires monitoring infrastructure that reaches the communities most exposed, innovation that lowers the barriers to access, data governance that converts evidence into decisions, and coordination mechanisms that align the institutions with the power to act. Clean air is not a technical problem with a single solution. It is a systems challenge that requires the right architecture to connect them all.