When we think of air pollution, we often think of Delhi, perhaps Beijing, or even Shanghai. Yet, our neighbourhoods from across the country as well as cities – from London and Paris to Toronto and New York – struggle with similar issues. Hence, the World Health Organisation (WHO) reports that 9 out of 10 people around the world breathe polluted air.
As humans, we contribute the most to air pollution by using energy to drive our vehicles, power our houses, run our data centers, and to travel. So much so that everything we use today was made at a factory that has contributed to air pollution.
Today, technology has become an enabler to help address air pollution. It can aid in better measurement, identify its sources, develop policies, forecast, predict, and apply logic to problem solving.
It can also provide elaborate opportunities for organisations and governments to optimise their operations and reduce their impact.
Thanks to Artificial Intelligence (AI), air pollution can now be addressed more effectively. AI can be leveraged to monitor, inform, influence, measure, detect, forecast, advise, reduce, and manage air pollution in our cities.
AI to measure air pollution
According to a report by Greenpeace, 22 out of 30 world’s most-polluted cities are in India. As per recommendations, India requires a minimum of 4,000 air monitoring stations to check air quality.
Currently, there are approximately 160 air monitoring stations in India. This implies we have less than 5 percent of the recommended number of air monitoring stations. This inadequacy is a hindrance to policy making, harming the potential to generate solutions.
A base dataset is needed for everything from city planning to new drug development. Air quality monitoring forms the base layer for a number of things, including human health.
Like Peter Drucker said: “If you cannot measure it, you cannot solve it”. AI can help solve this problem.
Accurate computation of air quality not only requires combining existing air monitoring infrastructure with satellite data but also factoring in human activities such as traffic, construction, garbage burning, industrial source apportionment, and population density. A feature engineered AI can be designed to utilise the above factors to accurately compute a geospatial interpolation of air quality data.
The advantage of such an approach is that it provides a distributed coverage of high spatiotemporal resolution in near real time.
To identify source, visibility, smog, and haze
In Delhi, the number of air monitoring stations have increased in the recent years. This has provided us with research-backed sources of air pollution within a few parts of the capital. By understanding about the sources of pollution, AI can be further used to track and predict the growth and reduction of air pollution.
For example, we could monitor whether an increase in industrial production is directly proportional to air pollution, or a decrease in vehicles is related to a reduction. These decisions could be evaluated by AI, allowing appropriate actions to be initiated.
AI can also help in modelling the chemical reactions between pollutants. Algorithms like Atmospheric Transport Modelling System (ATMoS) helps understand PM2.5 concentrations.
Additionally, there are advanced algorithms that help in understanding and predicting smog, haze, visibility, and observe meteorological interventions as well as manage air quality better.
AI and personal health
Air pollution in Indian urban environments has serious health and quality of life implications. A wide range of anthropogenic air pollution from multiple sources have contributed to an increase in the levels of air pollution, leading to a significant deterioration of ambient air quality.
AI can help track the current pollution baseline with respect to personal health indicators and advise people directly on avoiding exposure to dangerous levels of air pollution.
It can also provide actionable insights such as providing daily updates on air quality levels on their route to work and recommendations for safe spots to exercise, among others. AI can also track the improvement in the overall health index of a person who has followed these measures.
AI can help mitigate health risks and assist in drug development
Air pollution is linked to a number of health problems. This includes allergies, chronic respiratory disorders, pulmonary and cardiac issues, diabetes, and breast cancer. For enhanced treatment, these would require region specific clinical trials based on the patient’s exposure to the environment.
As results would differ, AI can help model such differences by looking at the samples and extrapolate it to the general population, developing region specific medicines and digital therapies among others to prepare humanity to handle the impact of air pollution.
The Environment Pollution (Prevention and Control) Agency or EPCA (a supreme court mandated panel), declared an emergency in 2019 on public health in the Delhi NCR region. Owing to the devastating and irreversible impact on our health, most health insurance plans now cover air pollution related diseases.
Many insurance providers have noticed this increase in air pollution through increased inflow of patients to hospitals. Both insurers and others must be prepared for health risks from air pollution linked to premature diseases.
A multitude of data and factors collected can help underwriters examine the risk and build support insure for such risks.
In summary, AI needs to be utilised across different sectors to cohesively address the massive global air pollution problem facing humanity.
For instance, AI can not only help forecast air pollution, but it can also help farmers improve nutrition through the improved understanding of nitrogen and anthropogenic elements in air and soil. When used ethically, AI can tackle issues that are plaguing the world.