Artificial intelligence improves air quality

A software program from Siemens is helping municipal authorities keep nitrogen oxide and fine dust concentrations under control.

Many cities around the world suffer from high concentrations of pollutants. City Air Quality Management (CyAM) software from Siemens uses artificial intelligence to calculate the pollution for many days in advance and determines the effectiveness of possible counter-measures.

More than half the world’s population now lives in cities And many of these cities have a problem: poor air quality. Harmful substances such as nitrogen oxides and fine dust regularly exceed legal limits, and residents – especially children and older people – suffer the health consequences, such as asthma and heart disease. This problem affects already disadvantaged groups in particular, since the air quality in poorer neighborhoods is often the worst. Reducing pollution is therefore not just an ecological issue but also a matter of social justice.

 

Many large cities have identified the problem and developed plans for improving air quality. But because the urban infrastructure cannot be changed quickly, many measures won’t take effect for another ten to 20 years. “But that’s far too late,” says Cathe Reams, Communications Director of Sustainability and Urban Infrastructure at Siemens. “We need to improve air quality in cities, and we need to do it as fast as possible.”

“We need to improve air quality in cities, and we need to do it as fast as possible.”
Cathe Reams, Communications Director of Sustainability and Urban Infrastructure at Siemens

The necessary technologies are already available today, such as alternative drives that produce no local emissions. Other options are sensors that can detect the level of harmful substances and provide an up-to-date overview of the air quality situation. This requires an intelligent tool that utilizes this information and suggests immediately effective measures for city administrations.

Precise forecasts of pollution levels

This goal is what City Air Quality Management (CyAM) from Siemens is intended to achieve. The system is based on smart software that combines the latest measurements of air pollutants, such as nitrogen oxides (NOx) and fine dust in particle sizes no larger than ten micrometers (PM10) or 2.5 micrometers (PM2.5), with the latest weather forecasts in order to derive predictions for the coming days. “CyAM can predict the concentrations of nitrogen oxides as well as PM10 and PM2.5 for the next three days with 90-percent accuracy” explains Reams. “The accuracy is approximately 80 percent for a five-day forecast period.”

CyAM can predict the concentrations of nitrogen oxides as well as PM10 and PM2.5 for the next three days with 90-percent accuracy.
Cathe Reams, Communications Director of Sustainability and Urban Infrastructure at Siemens

The precise forecasts are based on artificial intelligence, which learns how the air quality is likely to progress based on measured values and weather data from the past. The weather is important because moisture, sunlight, cloud cover, and temperature all affect pollution levels. But CyAM also distinguishes between working days and weekends and even considers recurring events such as trade fairs and sports events. A crucial factor is the availability of historic data, which should go back as far as one year or more for the best possible forecast quality.

CyAM has a dashboard where users can see the latest measured values from the city’s sensors as well as a forecast, either hourly or for several days in advance. A traffic light system makes it immediately clear how the values will develop. Green stands for good quality, while yellow and especially red show deterioration. But there’s more: The tool also makes suggestions on ways to quickly lower the pollutant concentration. “For example, the city administration can impose a temporary ban on driving diesel cars or trucks,” says Reams. “Or they can deploy electric buses in certain areas.”

Immediately evaluating the effectiveness of measures

CyAM also displays how effective these measures would be. They can be activated at the click of a mouse, after which the forecast immediately adapts to the new situation. Red and yellow lights turn into green areas. City administrations can thus make decisions based on objective data, don’t have to depend on assumptions, and can achieve their goals with the fewest restrictions on people. “We want to maintain a sense of proportion in all measures and, for example, impose a ban on diesel cars or trucks only on certain days or biannually,” Florian Ansgar Jaeger, project manager and air quality expert at Siemens, explains.

“We want to comply with the limits as quickly as possible. A tool like this can be very helpful.”
Peter Pluschke, head of the city of Nuremberg's environmental office

Several cities are already testing the CyAM tool. One of them is Nuremberg, which helped develop the tool by offering ideas and acting as a beta tester. The city has problems particularly with nitrogen oxide concentrations in its air. An annual average of 40 micrograms per cubic meter is permitted, but at 46 micrograms it was slightly above this level in 2018. Nuremberg has been using CyAM for a year now, primarily to evaluate potential future measures such as driving bans. “We want to comply with the limits as quickly as possible,” reports Peter Pluschke, head of the city’s environmental office. “A tool like this can be very helpful.”

The air quality forecast is currently being integrated into Nuremberg’s website. The city has also launched an RSS feed in multipliers such as retirement homes and public swimming pools, using CyAM to forecast ozone concentrations, among other things, and notify risk groups of dangerous pollution levels. Other cities around the world are also greatly interested in the tool. After all, people in metropolitan areas can’t wait for decades for better air quality.

09-25-2019

Picture credits: Getty Images /Siemens AG

Subscribe to our Newsletter

Stay up to date at all times: everything you need to know about electrification, automation, and digitalization.