What If AI Can Monitor the Air You Breathe?

Across India, this winter carries an unsettling undertone. With the festive season underway, along with the changing weather patterns, pollution levels have surged again across most of northern India. The National Capital Region, like every year, seems to be recording the worst numbers in its count of pollutants.  

Delhi provides the most vivid warning of what this trajectory leads to. After Diwali and the crop burning season, the capital once again slipped into the ‘Severe’ and ‘Severe Plus’ AQI categories. 

Several zones recorded PM2.5 levels above 450 and some touched 500. Visibility dropped, schools issued advisories and hospitals saw a jump in respiratory cases. This cycle has become so predictable that it is now considered part of Delhi’s seasonal rhythm. 

Even urban centres once considered relatively breathable, like Bengaluru, are now showing worrisome signs of deterioration. 

The city is experiencing a form of air stagnation, where polluted air stays trapped close to the surface, blocking sunlight and causing a deeper, unnatural dip in temperature. 

Bengaluru’s air quality this winter has seen a sharp decline, with early December recording some of the city’s highest pollution levels of the season. The city witnessed its worst spike on December 7, when the AQI hit 161, falling into the “unhealthy” category and marking the highest reading during the period over the last five years (graph below).

While pollution levels eased slightly after this peak, dropping back toward the 80–100 range, the early-winter surge highlights a worrying trend: Bengaluru, during winter, is increasingly experiencing significant pollution episodes that resemble those typically associated with northern Indian cities.

Air purifier sales across urban India reflect this rising fear. Retailers report a sharp increase in demand in the weeks following pollution spikes. Popular electronics retailer Croma reported a 30% increase in the year-on-year sales of air purifiers across its online and offline channels. 

According to IQAir’s 2024 assessment, India remains one of the world’s top five most polluted countries, with six of the 10 most polluted cities globally located here. Toxic air cuts more than three years off the average Indian’s life expectancy. 

Despite the scale of the crisis, India’s air monitoring infrastructure has not kept pace. Most cities still rely on only one or two official AQI stations, leaving vast gaps in understanding how pollutants move across complex urban microclimates.

This gap in measurement is one of India’s biggest blind spots. As Amit Banka, founder and CEO of WeNaturalists, says, “AI models using satellite-derived intelligence can virtually sense pollution in areas without physical sensors, creating a scalable virtual monitoring network.” WeNaturalists is a global ecosystem empowering professionals and organisations working with nature with digital solutions. 

Banka’s observation highlights a growing consensus among climate scientists that the problem in India is not the absence of data, but the absence of consistent, connected data capable of informing real-time governance.

How Can AI be Leveraged?

A 2024 study by Kuldeep Singh Rautela & Manish Kumar Goyal, published under Scientific Reports in the Nature journal, backs this urgency with hard evidence. Researchers used an AI/ML model, a convolutional autoencoder to forecast PM2.5 levels across India, even in regions with little to no ground monitoring. The model’s accuracy was unexpectedly high, with error margins under 10 µg/m³ and structural-similarity scores above 0.60, proving that AI can generate reliable pollution forecasts even when India’s sensor network falls short. 

Rautela & Goyal warn that India’s AQI infrastructure is too sparse to capture true exposure levels, and that AI-driven forecasting is no longer optional; it is essential if cities like Bengaluru want to avoid the “Delhi trajectory.”

“India’s official monitoring network is geographically limited,” says Banka, adding that AI can override these limitations by combining historical data with real-time satellite imagery.

Professor A Damodaran of IIM Bangalore, believes this is where India must act decisively. Ease of driving is being prioritised over ease of breathing, he told AIM.

It is perhaps the clearest articulation of the dysfunction at the heart of India’s pollution problem. Traffic police prioritise vehicle movement. Pollution boards prioritise emissions reporting. Neither system is designed to work with the other.

AI offers the glue that India’s institutions have long lacked. According to Damodaran, India’s biggest pollutant is still suspended particulate matter produced by older vehicles, autorickshaws and trucks. 

AI-powered camera-vision systems can already identify which types of vehicles pass through a junction and which ones are likely contributing disproportionately to particulate spikes. When correlated with traffic flow, such systems can recommend rerouting traffic in polluted corridors before the exposure becomes hazardous.

This shift is already visible in the startup ecosystem. Momentum Capital, which invests in climate and deep-tech ventures globally, sees AI as central to the next phase of environmental governance. Founder and managing partner Ankur Shrivastava says they are bullish on AI-led climate solutions that go beyond diagnostics and build mitigation tools, “from smart traffic-flow systems to industrial emission control.” 

Shrivastava points out that scaling these systems requires navigating difficult regulatory terrain. “Air-quality startups must navigate a complex policy ecosystem, which is why we prioritise founders who can build teams capable of operating across regulatory environments.”

The future of urban environmental intelligence in India looks increasingly agentic. Banka believes, “India is moving toward agentic AI systems that will not just alert officials to smog, but automatically trigger interventions like rerouting traffic, or activating suppression systems.” 

These systems can eventually power climate digital twins for major metros, giving decision makers the ability to run simulations that forecast the impact of electrifying buses, restricting construction dust or restructuring traffic before policies are implemented.

The Centre for Research on Energy and Clean Air analysis published a study last month that claimed that across 60% of Indian districts, “exposure to significant air pollution is not restricted to winter alone as is commonly believed.” The report noted that expanding the ground-monitoring network nationwide is essential. 

India can no longer treat air quality as an episodic, wintertime inconvenience. The pollution crisis is structural, shaped by rapid urbanisation, weak monitoring, fragmented governance and climate volatility.

The post What If AI Can Monitor the Air You Breathe? appeared first on Analytics India Magazine.

Generated by RSStT. The copyright belongs to the original author.

Source