As government authorities in China and elsewhere race to contain and treat the coronavirus making its way across the globe, technologies built around big data and artificial intelligence (AI) are increasingly being pushed to the forefront of the fight.
Specifically, many Chinese cities are busy building defensive measures using data and AI technologies against the epidemic.
The systems at work
Some Chinese tech firms have developed apps that check if users have taken the same flight or train as confirmed patients, according to a report on the AFP. By scraping data published by state media, the app attempts to backtrack and correlate movements of individuals to alert users if they are at risk of infection.
Elsewhere, at least one neighborhood committee responsible for an apartment complex of more than two thousand households in Beijing used flight and train data to keep track of their recent travels. Elsewhere, a man who traveled to Wuhan had police show up unannounced at his home in Nanjing to check his temperature after they identified him from travel data obtained from Wuhan.
With fever being one of the symptoms of flu, public transport hubs in China are also being fitted out with detection systems designed specifically to process large groups of commuters quickly.
At the Qinghe railway station, infrared and face detection technology from Baidu is used, which automatically photographs each person's face and their associated temperatures.
Should someone with an abnormal body temperature be detected, station staff is immediately alerted to conduct secondary checks – and can find them quickly due to their photos. The system checks more than 200 people a minute, which is significantly faster than the ones deployed at some airports.
One challenge revolves around tailoring these systems to detect temperatures with only the forehead exposed, given the orders for mandatory wearing of masks in public places in many parts of China. By working through the Chinese New Year holidays, AI firm Megvii had apparently succeeded in reprogramming its thermal scanners used at airports.
Humans still needed
Despite the role of data and AI, it must be noted that humans continue to play a vital role in this battle. For a start, data must be entered at some point for contact tracing to work. On that front, some neighborhoods in Beijing prompt residents to fill out their personal details and travel history on their smartphones by scanning a QR code.
A strong human component is also needed. In Singapore, seven teams made up of 10 people each work in two shifts from morning (8.30 am) till late at night (10 pm) on contact tracing, making phone calls to check for close contacts of confirmed coronavirus patients.
Patients are interviewed to determine their movements over the last 14 days. This includes friends they ate with, as well as places they have been to. Where information gaps exist, the police with their “investigative expertise” take over, according to a report from the Straits Times.
While the techniques used were left unmentioned, a deputy assistant commissioner alluded to the work as “painstaking data analysis and extensive legwork on the ground”. Likely, this entails the analysis of mobile phone location records and possibly police cameras located around the island city.
Work on other fronts continue apace. For instance, a group of researchers modeled human movement among cities in China from geolocated mobile phone records obtained from Tencent. By tracing the patterns of human mobility across China before the coronavirus was discovered, they found that the travel restrictions may have slowed its spread by a few days for critical countermeasures to be implemented.
To be clear, many of these systems are ad-hoc in nature or involve modifications of existing products before being rushed into service. Others, such as this visualization of the virus’s progress by researchers at John Hopkins, are more educational in nature.
Together, however, they illustrate the power of data and AI, and their inextricably roles in our society today.
Photo credit: iStockphoto/Dmytro Varavin