Are AI And Air Quality Monitoring Natural Bedfellows?

Artificial intelligence is rarely out of the headlines these days, particularly around matters like sustainability and the environment.

A recent report by the Climate Action Against Disinformation coalition warned AI has the potential to “turbocharge” climate disinformation.

In addition, the report warned AI systems require an enormous amount of energy and water, and consumption is expanding quickly, with estimates suggest a doubling in 5-10 years.

On the flip side, research by Pi Labs claims AI could reduce the annual carbon footprint of the built environment by 50% by 2030 and help create a greener world.

Away from the headlines, the growing air quality monitoring sector is one that is already making heavy use of AI and number-crunching algorithms to make sense of all the information being generated on a regular basis. Does this mean AI and air quality are a match made in data heaven?

Wiktor Warchalowski, the chief executive and co-founder of air quality specialists Airly, said in an interview that he sees AI as a friend.

Warchalowski told me he believes AI will not replace people, but he added people using AI will replace people who are not using it, as it speeds up processes and handles repetitive tasks.

He said AI will enable people to have more rewarding, better jobs and help retain and attract staff to work in technology-based sectors, like air quality monitoring.

“The world is changing and there is a need to attract top talent into a sector like ours, and if you are not using the latest technology and asking young people to do boring things, then they will give their notice, and leave” said Warchalowski.

Airly just launched a new AI tool which has been developed to revolutionise the work of air quality consultants in the U.K. and dramatically accelerate the production of air quality assessments in planning applications.

The platform utilises the power of AI to rapidly gather relevant data from both public and private sector online resources.

It also includes monitoring data from local authorities’ annual status reports, as well as background concentrations, human receptors and ecological receptors.

“Our vision is to revolutionise environmental consulting by building digital tools that will like boost productivity,” said Warchalowski.

“We decided to start with air quality assessment for planning, because we saw like a lot of repetitive tasks in that in that area.”

Professor Shweta Singh, from the University of Warwick’s Business School, said in an email calculating air pollution in one place is not too complicated, but calculating it continuously and taking into account hundreds of variables is highly complex and lends itself naturally to outsourcing to AI.

“To go even further and try to predict the environmental situation or level of air quality whilst including theoretical data is even more complex, and so naturally AI can be useful here too,” added Professor Singh.

But Martin Butler, a professor at Vlerick Business School said in an email he believes AI is not particularly well suited for air quality monitoring.

Professor Butler said monitoring air quality creates a continuously stream of data, requiring significant computing power for capture, storage, and processing.

He added air quality alerts are typically issued when data approaches or breaches predefined thresholds or recognising predefined trends.

“This does not require AI and uses well-known and established automation algorithms that have been used in the industry for the best part of three decades,” said Professor Butler.

But he added AI could be used to visualise air quality at a level more granular than what is measured through interpolation and using other data sources to create rich datasets for visualisation and analysis.