COVID-19: Guj-based startup develops monitoring system for quarantine enforcement using AI

The Coronavirus pandemic has put hundreds of thousands of infected people across the world into strict quarantine, all in an effort to contain the spread of the deadly virus. Authorities in India, which is in the midst of an unprecedented lockdown, are scrambling to detect people infected with the virus and put them in isolation either in their homes or special quarantine centres in hospital and other places.

With a rising number of cases, the stretch on authorities to ensure that people with symptoms (even mild ones) remain in quarantine is increasingly getting tighter. However, like every conventional problem people face nowadays, the solution is being found in technology.

To help authorities meet the challenge of ensuring those in quarantine remain so, a Gujarat-based tech startup has developed an artificial intelligence-based system to allow monitoring of such individuals through their smartphones and leveraging GPS technology to give real-time information to authorities of their movements.

READ | IIT-Guwahati Students Develop Drones To Sanitize Large Areas; Claim '1 Drone = 20 Workers'

Using artificial intelligence

Faceometrics is a GPS-based facial recognition system that allows authorities to keep a constant check on every quarantined person digitally, decreasing the manual and time-consuming task. Incubated at GLS Leaf and Product Development support by NASSCOM COE – IIT Gandhinagar, the company — Indolytics — has sought to tie-up with the Ahmedabad Municipal Corporation and the Rajkot Municipal Corporation to help the civic bodies monitor under-quarantine people.

Monitoring system

The user, through the Faceometrics app, has to periodically upload a photo of them inside their home or the quarantine centre, which will reflect on the system of the authorities. An automatic live GPS location matches the photo's location tag against home GPS. If the user fails to log their attendance in the stipulated time, an alert is raised in the system of the authorities. 

If the user is 100 meters away from their quarantine room, an alert is raised. A notification of the same is also sent to the user automatically. Using AI, the system calculates the risk ratio of the person based on the number of violations occur. The audit, done over AI, is then accessed by the authorities. This touch-free automated inspection decreases the chance of viral transmission.

Speaking to Republicworld, Indolytics co-founder Ankit Dave said that the system also recognizes cheat attempts by users and identifies fake photos or videos. If an attempt is made to forge image, a notification to authorities is sent.

READ | Indian Engineering Student Develops A ₹500 Device To Ensure Social Distancing

User data protected

When asked about concern over privacy and data protection, Dave said, "We are working on an SOP (Standard Operating Procedure) with the authorities. We have also ensured user data privacy is protected. The app does not require contact information, just access to the location. The user's image is not saved on the server, it's just a landmark's tag that's fed in the system." He also said that the company's servers are based in India.

Indolytics has also approached the World Health Organisation (WHO) to partner in quarantine efforts and plans to do the same in metros like Mumbai, Delhi and Bengaluru, Dave further said.

READ | 'ThankYou Coronavirus Helpers' - Google Honours Public Health Workers And Researchers

READ | Coronavirus LIVE Updates: Centre Mulls Extending Lockdown Post-April 14; Cases At 5274