Contact Tracing; Yes it’s a difficult task to do and Sri Lanka’s Epidemiology Unit, Ministry of Defense, Central Intelligence Department and others are trying hard to capture these information.
Let’s say, if we build a Corona Tracking App that can track location data, calls, SMS and everything you’ve got. But unfortunately why someone should install it on the Day 1 of the decease identification to the world?
In countries like China and South Korea,they use tracking data (WeChat, CCTV Surveillance Data) to monitor the people who has been identified positive with the COVID-19. But we; Sri Lankans don’t have some high end tech platforms where all the CCTV connected to one Central Database with high ML capabilities, our own social network or even a very common app as WeChat on our own.
At the time of writing this post, in Sri Lanka there are 117 corona positive patients identified and we are still trying to trace all the contacts of them. And epid.gov.lk is already doing an impressive job trying to find everything they can regarding this matter as shown in their PDF reports.
As an example we can take the “Jaffna — Philadelphia Missionary Church Incident”. All the people who attended to the service (around 150 people) were advised to self-quarantine on 21st of March (after more than five days since the prayer service). And there are no any tracking info we can check for the people who didn’t attend to the prayer, but got the disease due to the carriers of virus who came for the prayer.
Google has the capability to gather detailed data of location history, from each and every user of Android and through some Google apps like Maps. But, for countries like ours, the Google cannot track who are the COVID-19 positive patients and who should be self-quarantined hence there are no link with social profiles when identifying a person.
Therefore what we can do in countries like ours is, we can build an app where the patients can input their location history data and with that we can identify their paths to find the Geo risk locations. And also anyone can use that app to find whether they have been to such places unknowingly by inserting their location history data to the app as well.
For a specific day, google knows all the locations where a person went with the timing as below
and with the Location History we can intercept the corona carrier path with others’ paths.
With that we can easily get all the Geo Risk locations in to one place and identify the people who got exposed to these locations.
I made a prototype to get the location list extracted for a printable PDF report from location history. And I’m willing to develop this app more, into identifying the present risk factors and to help people to find and guide to self isolate.
#What it Does
With the research I did, I have found only the Google Fit Location History tracking API for the use-case. But that doesn’t generate the correct data I need from iOS users who also engaged with Google Maps.
I came up with a simple website that allows anyone to download your Google Location History data (assuming you left it active on your account) and list them by date, time and address.
Google Takeout — as the tagline says It’s Your account, your data.
#How I Build — TL:DR
With some Java Script coding, HTML and fork of git, I came up with a Location extractor for Google Location History export.
Let’s say you are infected with COVID-19; how do you easily recall the locations you visit and tell others (Health Officers or your friends), these are the places that I have been visited and you must mark these as Geo Risk Locations?
#How to Run
- Step 1 : Navigate to #CoronaTrace
- Step 2 : Download your Location History from Google Takeoutusing the steps shown below
- Step 3 : Get the Selected months JSON from the ZIP file
- Step 4 : Select the File from the File Chooser (Not the ZIP file you have to extract and upload the Specific month’s JSON)
- Step 5 : Now You can see the Table Processed as below and Print As PDF
First, purpose limitations are necessary to ensure that all parties can only collect and share data for the purposes of combating COVID-19, nothing else.
Therefore the privacy issues of accessing private location data was the main challenge I faced. But with the Google Takeout people can easily export their location history, and I made my website as they can visualize that data without tracing them to the server.
#What’s Next for #CoronaTrace
There are many steps ahead
- Data Visualization and Heat-map generation from Location History JSON.
- COVID-19 Positive patient Tagging with Locations.
- Trace if you are vulnerable.(Main point of the project)
- User Data capturing through API (This will be a hard part)
#Try out Links
- Dashboard — http://coronaupdate.nisal.me/
- Tech Behind Dashboard = https://nisal.me/blog/2020/04/tech-behind-covid-updates-dashboard/
- Patient Update — https://coronaupdate.nisal.me/patient.html
- CoronaTrace — https://coronaupdate.nisal.me/trace/
- Locatin History Survey — https://nisalme.typeform.com/to/JsTyTY