Linda Poon is a staff writer at CityLab covering science and urban technology, including smart cities and climate change. She previously covered global health and development for NPR’s Goats and Soda blog.
IBM says its artificially intelligent machine can forecast air pollution up to 72 hours in advance.
IBM already conquered Jeopardy with its supercomputer. Now the tech giant, which has become one of the leading pursuers of artificial intelligence, wants to use that technology to fight Beijing’s air pollution crisis.
The company’s researchers are currently testing a computer system that can predict the city’s pollution levels 72 hours in advance. It will be 30 percent more precise than conventional methods of predicting air quality, a leading researcher told Technology Review. Using old data from the Beijing Environmental Protection Bureau, IBM is training the machine to forecast future patterns while taking into account changes in industrial activity, traffic congestion, and the weather—a process dubbed “adaptive machine learning.”
Over time, researchers hope the system can make accurate predictions up to 10 days in advance. That will help Beijing generate more specific measures, Tech Review reports, such as putting a cap on the number of cars on the streets on days when the air is expected to be particularly dangerous. It’s all part of Green Horizon, IBM’s 10-year initiative to help China manage its air quality and energy systems, as well as protect the health of citizens.
The smog in China is among the worst in the world, with the average level of pollutant particles more than twice the level recommended by the World Health Organization. The poor air quality not only threatens the country’s tourism industry but also kills approximately 4,000 people a day, according to a recent study out of University of California, Berkeley.
For its part, China has committed to curbing air pollution 25 percent by 2017. Cities like Beijing and Tianjin have enacted tougher regulations, such as limiting the number of new car registrations and closing down factories and mills in the most polluted provinces. For example, the province of Hebei, which is home to seven of China’s 10 most polluted cities, reported that it has closed down roughly 18,000 factories since 2013 in an effort to clean the air.
Still, the country is years from bringing its pollution down to safe levels. IBM could help move the process along by encouraging drastic yet temporary measures. But in the long run, the new system would have to trigger a significant behavioral change in the 1.4 billion people living in the world’s largest economy, where ceasing factory production comes at a heavy cost to local economies and where companies and entrepreneurs alike have found ways to capitalize on the dirty air.