Surveillance, an anti-pollution weapon

 

 

In the future, China will shut down a factory before it even pollutes — or so it hopes, as it deploys big data in the fight against bad air.

In Beijing’s environmental bureau, a team of engineers tend to giant mainframe computers that keep a watchful eye on the city’s pollution.

Using everything from factories’ infrared profiles to social media posts, the machines can call up three-day pollution forecasts with resolution of up to one kilometre squared and detect trends up to 10 days out.

Drones, satellites, sensors

The computer program, developed by IBM, is one of several high-tech measures, ranging from drones and satellites to remote sensors, that China is deploying to deal with its chronic pollution.

It seeks to solve an incongruous reality: In a country where security cameras are ubiquitous and Communist authorities operate a vast public surveillance system, accurate information about pollution remains scarce — even to officials.

As a result, Beijing and its neighbouring provinces “can’t coordinate joint defence and joint control” of their anti-smog efforts, leaving rogue companies to “secretly discharge and secretly dump”, said Chen Long, chief executive officer of Encanwell, which develops air quality monitoring and early warning systems.

The company is trying to achieve total pollution awareness: the ability to know, with perfect accuracy, where haze comes from and use that information to predict and preempt its future sources.

China has found itself in a double bind in the face of a relentless assault from bad air that put the capital on its first-ever air quality red alert this month.

Improving prediction

Choking pollution descended on Beijing twice in the past two weeks, and the country’s meteorological bureau expects it may come twice more before the month is out.

Improving prediction capabilities through increased monitoring is a key part of the plan to trade in the country’s anti-pollution sledgehammer for a scalpel.

In July, leadership in Beijing vowed to establish a national network for detecting pollution that will incorporate multiple technologies on land, in the air and space.

So far, the IBM software has achieved only 75 per cent accuracy for its 10-day forecasts, developers say. — AFP