CAWIL.AI Solutions, an industry-agnostic artificial intelligence (AI) social enterprise, is in talks with the local government units (LGUs) of Calapan City and Quezon City for pilot testing a contact tracing system that will help offices and public areas enforce health protocols—such as physical distancing and the wearing of face masks and face shields—to prevent the spread of coronavirus disease 2019 (COVID-19). 

“The requirement for LGUs to be able to adapt to the system is a working CCTV camera or an Internet Protocol (IP) camera, which is already mandated in every local government,” said Cherry B. Murillon, CAWIL.AI’s founder and lead innovator. Other requirements are internet access and subscription to their system, which starts at P500,000 for one year, and covers the installation of two cameras per location. 

“We hope that all establishments—and not just LGUs—could adapt the automated system without compromising data privacy,” she added. “It is important to really know what data the government collects and how it is being used.”

The artificial intelligence (AI) and data analytics-powered system was developed using computer vision technology from pre-installed closed-circuit television (CCTV) cameras. Using data gathered from CCTV videos, the system can determine whether a location is operating at the capacity required for physical distancing, usually 50% of actual capacity. 

A COVID-19 map also provides information on the proximity of people, which can then be printed out or generated for reference on possible overcrowding. The system was tested in SM City North EDSA’s Annex, a mall known for its high density of human traffic.

“The data we gather is the number of people that pass through the entries and exits of a public establishment,” said Ms. Murillon. “We can alert the establishment if the number of people has exceeded the physical distancing requirement.” 

She added that the system does not breach data privacy because it does not collect personal data, and that it is more useful than sharing information on health declaration forms with no actionable input. “Contact tracing’s goal is prevention. Collecting names and contact numbers will not prevent the spread of the virus, regardless if it’s paper-based or through Quick Response (QR) codes.”

Data is protected in the cloud and is encrypted depending on the level of access the establishments prefer. All data will be turned over to the establishments at the end of each subscription year to prevent fraud. 

CAWIL.AI received a grant from the Innovation for Social Impact Partnership’s Innovative Solutions Grants Facility. — Patricia B. Mirasol