Use of big data, AI to curb Covid-19

 News that matter in Papua New Guinea

Response to Covid-19 in Taiwan

Use of big data, AI to curb Covid-19

PORT MORESBY: In the first installment of our series on how we can try to restore normalcy by learning from the Taiwan experience, we now start by looking at how the Taiwanese used big data and Artificial Intelligence (AI) to tackle the Coronavirus (Covid-19) pandemic. 

The earlier report was this: https://pngcybermonitor.blogspot.com/2021/07/fighting-covid-19-taiwans-successful.html (Fighting Covid-19: Taiwan’s ‘successful’ experience)

Here is Part 1:

Leveraging on big data for face mask supply

Taiwan’s pioneering name-based rationing system for face masks:

On Jan 31, 2020, the Taiwan Government announced that it would requisition all face mask factories in order to centrally manage the distribution and production of face masks in the country.

A total of 2.8 million face masks were released daily for sale to the public through convenience stores, drug stores and pharmacies. However, stores received relatively small face mask shipments, and these shipments arrived at different times each day.

As a result, many people complained they were not able to buy face masks, leading to panic buying and hoarding, thus impeding epidemic prevention efforts.

Following careful deliberations, the Government launched the name-based rationing system for face masks (version 1.0) on Feb 6, which allowed people to buy face masks at pharmacies contracted by the National Health Insurance (NHI) Administration with their NHI cards.

Prior to the implementation of this system, an engineer from the private sector named Wu Chan-wei developed an app showing user-provided data on face mask availability at convenience stores, which proved to be very popular.

Digital Minister Audrey Tang, therefore, recruited Chan-wei and other private sector engineers through the online gov community, and together they developed an app within 48 hours that showed inventory levels at pharmacies.

This new app was launched simultaneously with the name-based rationing system in a bid to relieve people’s anxiety with regard to face mask purchases.

Tang facilitated communication between the Government and private sector, obtaining open data for app development as well as user feedback. NHI administration data was renewed with even greater frequency, going from updates every 30 minutes to updates every 30 seconds.

Online ordering added to the name-based rationing system:

Even after the implementation of the name-based rationing system, demand was still not fully met. Long queues formed outside pharmacies because there were only 6,026 NHI-contracted pharmacies and local public health centres nationwide. Manpower was also limited.

The Government continued to discuss possible solutions, such as once again including convenience stores with sufficient manpower and large geographical coverage among distribution points and enabling people to order face masks online.

The Government realised that convenience stores in remote areas often had a surplus of face masks and that office workers and students who needed face masks for their daily commute via public transportation were not able to buy them because they lacked the time to line up outside pharmacies.

In order to distribute face masks more evenly, the Government worked with the private sector and launched the updated system (version 2.0) on March 12, about one month after the initial launch.

In addition to face mask purchases at NHI-contracted pharmacies and local public health centres, the updated system gave people the option to order face masks online by logging into the eMask website with their NHI card or Citizen Digital Certificate.

They could also use the NHI app to order face masks on their mobile phones.

Following the addition of this new function, the total download count for the NHI app increased to more than 7.3 million, making it the most downloaded app in Taiwan in 2020.

People were given one week to go online and order face masks, which they could then pick up at their designated convenience store.

Once an online order had been placed, a text message was sent to the person making the order showing payment information. This information could also be found online after logging into the system.

Payment options included ATM transfer, online banking transfer, and credit card. Those who made the payment within the given deadline could then pick up their face masks at their designated convenience store with the text message and their NHI card, National ID Card, or driver’s licence (any of these three).

The upgraded version of the name-based rationing system gained widespread praise, mainly because the addition of convenience stores as distribution points increased service coverage extensively.

The four major convenience store chains in Taiwan operate more than 10,000 stores nationwide. Convenience stores are also often closer to people’s homes, and make it easier for people to pick up their face masks during their lunch break at work. Convenience was thus greatly enhanced.

In addition, the upgraded system reduced queuing at pharmacies. For convenience stores, distributing face masks was not a heavy burden on their personnel and could help attract more customers, thus boosting sales.

Many convenience stores were therefore glad to participate and even came up with promotional sales, such as two cups of coffee for the price of one. People no longer had to squeeze time out of their schedule just to line up and buy masks.

They could order online and pick up face masks whenever they wished, and even enjoyed promotions at convenience stores. People recommended the online ordering mechanism to one another, and usage increased considerably. It benefited all parties involved, including the Government, the public, and convenience stores.

Third iteration of the name-based face mask rationing system:

With the increase in face mask production, the Government sought to resolve the problem of elderly people still lining up outside pharmacies. Therefore, the third iteration (version 3.0) of the name-based face mask rationing system was officially launched on April 30, 2020, following careful planning. Restrictions on face mask purchases at pharmacies related to National ID Card or Resident Certificate numbers ending in odd or even numbers were cancelled.

When picking up previously ordered face masks at convenience stores, people could now also order face masks for the next batch at the same time by simply inserting their NHI cards into kiosk machines at these stores, which enabled people to complete their order within one minute.

The implementation of this new upgrade aimed to reduce the amount of time people spend standing in line for face masks, alleviate the burden on pharmacies, and help convenience stores boost sales.

With regard to the issue of personal information, Tang reassured the public via Facebook that the convenience store kiosks only verify National ID numbers and check the validity of NHI cards; they cannot access any other information.

To help other countries combat Covid-19, the Central Epidemic Command Center (CECC) launched a campaign titled “Protect Taiwan, Help the World” on April 27, allowing the public to donate their masks to other countries in need through the NHI app. This campaign was enthusiastically received: about 560,000 people responded in the first two weeks, donating around 4.4 million face masks.

On June 1, after Taiwan had built a sufficient stockpile, the Government lifted its ban on mask exports and removed restrictions on the number of masks that the average citizen could send overseas.

Photo: Alberto Buzzola/LightRocket/Getty ImagesFebruary 23, 2020 photo in Taipei shows people wearing surgical masks in a temple.

Use of technology to upgrade quarantine and epidemic containment efforts.

Building a multi-layered epidemic containment network:

The Taiwan government’s quarantine approach consists of border and home control. As the number of confirmed cases continued to rise rapidly around the globe, it became essential that inbound travelers fill out entry forms truthfully and in detail. On Feb 16, 2020, the Government implemented the Entry Quarantine System.

Inbound travelers were requested to scan a QR code prior to their flight or upon arrival to fill out a health declaration and other information. This helped speed up customs clearance and ensured accuracy of information provided.

Traveler information was then integrated into the Home Quarantine Information System, which tracks people under quarantine or isolation for 14 days, and a digital fence system, which uses phone signals to monitor the location of people under home quarantine or isolation.

If people under quarantine or isolation leave their prescribed location, they immediately receive a warning via text message. Civil affairs authorities, health agencies, and local police are also notified in order to keep track of these individuals.

Those who leave their location without permission are subject to large fines, as are people who conceal symptoms associated with Covid-19. Individuals who complete their quarantine or isolation in accordance with regulations receive a compensation of NT$1,000 for each day spent in quarantine or isolation.

As of August 2020, fines totaling NT$116 million had been issued to those who violated the rules.

As numerous events were held across Taiwan on Dec 31 to celebrate the New Year, the CECC introduced version 2.0 of the digital fence system to detect individuals under quarantine or isolation or in self-health management who might be in the vicinity of large-scale events in violation of regulations.

It emphasised that personal data thus collected was deleted after 28 days to protect the privacy of individuals. As such, the digital fence system not only monitors people to prevent disease transmission but also protects their personal privacy.

More than 10 countries have inquired about the technologies that Taiwan employed.

To monitor the health status of an increasing number of people undergoing quarantine after returning from abroad, the Government developed an automatic operation system called Disease Containment Expert (DCE).

The system was launched on April 3, 2020, allowing those under quarantine or isolation to report their health condition as well as receive relevant information, thereby alleviating the burden on frontline personnel.

In addition, the DCE sends related guidelines to individuals on the day they complete their quarantine or isolation and two days prior to that, reminding them to conduct another seven days of self-health management.

The DCE initially offered only Chinese and English-language interfaces, but added four other language options - Indonesian, Filipino, Thai, and Vietnamese - in late December to facilitate the submission of health status reports by foreign nationals.

The government is also employing NHI big data. By collaborating with the National Immigration Agency and Taiwan Centres for Disease Control, the NHI administration can crosscheck the list of citizens returning to Taiwan from abroad with their NHI records.

Doctors on the frontline can check their patients’ travel history for the past 14 days and thereby help curb the spread of the disease. Using the NIA and CDC databases, the government can conduct big data analysis to identify Covid-19 cases through travel histories and reported symptoms.

 

It took only 14 days for Academia Sinica to successfully synthesised the first group of monoclonal antibodies capable of recognising the new coronavirus protein.

Epidemic containment made more precise with AI and big data:

As for the use of AI, then vice -premier Chen Chi-mai posted an article on his Facebook page on March 26, 2020, titled “Using AI and Big Data to Accurately Tackle the Covid-19 Outbreak”.

The article indicated that in late January the Executive Yuan had formed a big data task force to explore how to widely use smart technology in the fight against the outbreak, inviting experts to join the undertaking.

Some of their early ideas included:

Ø USING using AI to assist clinical physicians with diagnosing Covid-19 by implementing big data and integrating outbreak surveys, clinical and laboratory testing, and virus analysis to make epidemiological predictions;

Ø USING AI to help enforce and manage home quarantine and isolation measures; and

Ø DRUG repurposing platform: using AI, doctors have identified effective treatment drugs and are now continuing to analyse and investigate more drugs in laboratories.

Chen worked with experts at the CDC, the Executive Yuan’s Department of Cyber Security, the NHI, and National Taiwan University’s College of Public Health to compose an article explaining how the Taiwan Model applied big data analytics to prevent the further spread of Covid-19. The article was published in the Journal of Medical Internet Research on May 6.

Effectively managing high-risk groups and conducting community-based prevention:

In order to prevent cluster infections resulting from long queues outside public places, with people waiting for staff to take their temperature with handheld forehead thermometers, Microsoft Taiwan, MediaTek,and other companies have been using Artificial Internet of Things (AIoT) to detect face masks and body temperature and develop remote working tools.

They have come up with a one-stop station for mask detection and infrared temperature reading, using cameras and infrared thermal imaging technology to simultaneously determine whether people are wearing masks and whether their body temperature is within a normal range.

In order to track down suspected or confirmed cases, National Yang-Ming University’s Digital Medicine Centre developed a platform that monitors people for signs of Covid-19 infection.

The centre is providing it free of charge to domestic companies and health care institutions to help them monitor the health conditions of people in their organisations around the clock to effectively curb the outbreak. This cloud-based platform includes features such as body temperature monitoring, GPS tracking, contact history tracing, and automatic warnings.

Health management personnel can log into the backend system and check on the health condition of their members. AI can be used to classify risk levels and provide recommendations for epidemic prevention measures. The platform can be operated in multiple languages, including English, Japanese, Korean, Indonesian, Vietnamese, and Thai, to meet the needs of foreign nationals.