In Focus - Issue 32 (Spring 2020)

Quarantine tracking seeks to mitigate outbreak n order to enforce Hong Kong’s home quarantine order, simply using a GPS signal is not e ective because such a signal is o en weakened or blocked by the city’s numerous high-rises. Many buildings are also connected directly to shopping malls, without any signi cant GPS detection, creating a loophole for con nees to leave their designated areas. However, Prof. Gary CHAN, Computer Science and Engineering, had just the answer: a cutting-edge geo-fencing technology which detects whether a con nee is inside or outside a prede ned area by fusing, understanding, and learning various existing signals in the area. Prof. Chan has conducted research and development on fusion-based signal learning technologies for many years. One application is indoor positioning, with his research team devising and deploying, via his company Compathnion Technology Ltd., a smart location nder to assist visitors in shopping malls or hospitals navigate their way around. He now saw how the signal fusion concept could be further developed, adapted, and applied as a user-friendly, e cient, and cost-e ective way for quarantine monitoring. Prof. Chan’s team of researchers and engineers got to work, coming up with an automated geo-fencing technology they called “Signature Home”. Geo-fencing involves setting up a virtual perimeter for a physical area. The idea was then deployed by Compathnion as a mobile app called StayHomeSafe. Paired with an electronic Bluetooth wristband worn by the person quarantined, the app could detect whether the quarantine order was being complied with, and alert the authorities if not. The Signature Home technology is based on the concept that the collective signal values within a certain area are unique to that place, forming its “signature”. It works as follows: Following the March introduction of a -day mandatory quarantine period for Hong Kong people or visitors coming into the city, the Hong Kong government adopted the StayHomeSafe app for enforcement purposes. Such geo-fencing technology also achieves high user privacy because the system does not track nor know a user’s exact position, according to Prof. Chan. It only detects whether a person is inside or outside his/her home. To further enhance location privacy, the system can be designed to only inform the authorities when the person is detected outside their designated area. Mr. Arthur CHAN ( BEng in Computer Science), Director and CEO of Compathnion, added that the geo-fencing technology meant the app “can more e ectively safeguard public health”. I Cover Story 18 IN FOCUS The technology collects the various signals in a speci c home environment – Wi-Fi, Bluetooth, cellular, for example – and uses them as the home’s signature. If the collected signal values at a certain time deviate from the signature, it is likely the person has le the designated area. By using machine learning and data analytics techniques, the technology can also intelligently adapt to the evolving home environment to achieve accurate monitoring. Prof. Gary Chan’s mobile app, paired with an electronic Bluetooth wristband, was adopted by the Hong Kong government for quarantine tracking of inbound air travelers from March 2020.

RkJQdWJsaXNoZXIy NDk5Njg=