search

ehidra / contact-tracer

A prototype contact tracer app for COVID-19 pandemic response
MIT License
50 stars 19 forks source link

Pulling together a team #6

Closed baradhili closed 4 years ago

baradhili commented 4 years ago

Hi Guys, I've been working away and can now bring Public health support from the Scandinavian countries and I can also release info about this project to >500 epidemiologists all connected to the Norwegian Institute for Public health...

But I'm a bit worried.. its been really quiet here, so are things continuing here or should I take it elsewhere?

If its going to be here we need to roll up our sleeves and work out what we need and what features for the MVP.. this can go really fast if we want - we NEED to go really fast since Covid isn't waiting..

baradhili commented 4 years ago

pretty much the top priority is to spawn out a list of features needed.. modelling on TraceTogether is a good idea and I'm still working on them to release source.. ,

wickywaka commented 4 years ago

I will share the link in a Pakistani covid-19 group. Lets hope we will attract some devs from there.

8lenka8 commented 4 years ago

Hello,

I'd like to join as a data scientist. Our team have just done a contact tracing app as part of a GSA / Galileo location focused hackathon.

We used a bluetooth dataset in a graph structure for defining contacts. Location could be switched on in-app voluntarily by a user who is alerted to be at risk and used in aggregated for in dashboards for prioritising disinfection services.

I can help with data considerations, algorithms etc.

One thing to consider is time. If we take bluetooth readings every minute, perhaps 'contact' should be >2-5 minutes, not anyone walking by on the street. Then dataset size... 50 users over a month could generate 1/2 million data points. So preferably we want to aggregate contact data on device to day level... Anonymity, masking MAC ids which are PPI data... looks like Singapore solution is handling it very well. Alerting for every confirmed case contact will alert almost everyone. We can calculate simple probability e.g. if have several encounters or longer contacts the probability is higher.

wickywaka commented 4 years ago

Did you guys used any qr code based contact confirmation schemes? It can be useful in countries where not everyone has a smart phone. One person scan the other partys qr code before meeting them. That can be recorded in the smartphone and uploaded or stored.

On Tue, Mar 24, 2020, 18:05 8lenka8 notifications@github.com wrote:

Hello,

I'd like to join as a data scientist. Our team have just done a contact tracing app as part of a GSA / Galileo location focused hackathon.

We used a bluetooth dataset in a graph structure for defining contacts. Location could be switched on in-app voluntarily by a user who is alerted to be at risk and used in aggregated for in dashboards for prioritising disinfection services.

I can help with data considerations, algorithms etc.

One thing to consider is time. If we take bluetooth readings every minute, perhaps 'contact' should be >2-5 minutes, not anyone walking by on the street. Then dataset size... 50 users over a month could generate 1/2 million data points. So preferably we want to aggregate contact data on device to day level... Anonymity, masking MAC ids which are PPI data... looks like Singapore solution is handling it very well. Alerting for every confirmed case contact will alert almost everyone. We can calculate simple probability e.g. if have several encounters or longer contacts the probability is higher.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/memiah/contact-tracer/issues/6#issuecomment-603373554, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABVDYA5542YDGTRVQ6KHCDDRJDR43ANCNFSM4LSORDQQ .

8lenka8 commented 4 years ago

I have worked in a mobile/remote payments company Pomelo Pay which uses QR a lot, especially in Asia.

8lenka8 commented 4 years ago

Yes it seems like a good idea, but what is the scope of this project? UK or global? I can put you in touch with the international QR payments dev team in Pomelo Pay and they may be able to help. Where are you based? Is there a slack channel here so we could connect?

wickywaka commented 4 years ago

May project maintainers can answer these questions.

On Tue, Mar 24, 2020, 19:41 8lenka8 notifications@github.com wrote:

Yes it seems like a good idea, but what is the scope of this project? UK or global? I can put you in touch with the international QR payments dev team in Pomelo Pay and they may be able to help. Where are you based? Is there a slack channel here so we could connect?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/memiah/contact-tracer/issues/6#issuecomment-603437245, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABVDYA3SPEA64W36JIRKUD3RJD5EVANCNFSM4LSORDQQ .

8lenka8 commented 4 years ago

but how do you sugest to use qr codes without smart phone?

8lenka8 commented 4 years ago

to scan users' codes from paper by third party?

wickywaka commented 4 years ago

No. Only one phone would be needed for registering a contact. For example, a person going to a coffee shop can scan the coffee shop qr code. If other peoples who visited the coffee shop around that specific time and they turns out to be infected, we can have some potential infected people. I don't know if everyone would like to turn on their bluetooths. I personally keep my wifi and bluetooth off.

On Tue, Mar 24, 2020, 19:45 8lenka8 notifications@github.com wrote:

to scan users' codes from paper by third party?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/memiah/contact-tracer/issues/6#issuecomment-603439322, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABVDYAYH5ERJJJAKCZ724VDRJD5TZANCNFSM4LSORDQQ .

wickywaka commented 4 years ago

May be we can convince small businesses owners to convince people to use the app, if they want to be informed about any possible infections from that shop/caffee. I saw lots of QR codes being used in china. We can look if we can find something about how they used it and if those can be applied here.

On Tue, Mar 24, 2020, 20:23 Waqar Rashid waqarrashid33@gmail.com wrote:

No. Only one phone would be needed for registering a contact. For example, a person going to a coffee shop can scan the coffee shop qr code. If other peoples who visited the coffee shop around that specific time and they turns out to be infected, we can have some potential infected people. I don't know if everyone would like to turn on their bluetooths. I personally keep my wifi and bluetooth off.

On Tue, Mar 24, 2020, 19:45 8lenka8 notifications@github.com wrote:

to scan users' codes from paper by third party?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/memiah/contact-tracer/issues/6#issuecomment-603439322, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABVDYAYH5ERJJJAKCZ724VDRJD5TZANCNFSM4LSORDQQ .

wickywaka commented 4 years ago

Or given the life of virus on phones surface, I don't like carrying my phone outside of my home. So would be a good idea to have something that can register a contact without everyone having a phone. If govts adopt it, they can use it as visitors list. I saw some QR code being used in a Singaporean university for classes. It might solve the proproblem of figuring out who attended a controlled social event.

On Tue, Mar 24, 2020, 20:25 Waqar Rashid waqarrashid33@gmail.com wrote:

May be we can convince small businesses owners to convince people to use the app, if they want to be informed about any possible infections from that shop/caffee. I saw lots of QR codes being used in china. We can look if we can find something about how they used it and if those can be applied here.

On Tue, Mar 24, 2020, 20:23 Waqar Rashid waqarrashid33@gmail.com wrote:

No. Only one phone would be needed for registering a contact. For example, a person going to a coffee shop can scan the coffee shop qr code. If other peoples who visited the coffee shop around that specific time and they turns out to be infected, we can have some potential infected people. I don't know if everyone would like to turn on their bluetooths. I personally keep my wifi and bluetooth off.

On Tue, Mar 24, 2020, 19:45 8lenka8 notifications@github.com wrote:

to scan users' codes from paper by third party?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/memiah/contact-tracer/issues/6#issuecomment-603439322, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABVDYAYH5ERJJJAKCZ724VDRJD5TZANCNFSM4LSORDQQ .

wickywaka commented 4 years ago

Programming wise, it can be simply adding a feature that will allow for a contact to be registered by scanning a QR code.

On Tue, Mar 24, 2020, 20:28 Waqar Rashid waqarrashid33@gmail.com wrote:

Or given the life of virus on phones surface, I don't like carrying my phone outside of my home. So would be a good idea to have something that can register a contact without everyone having a phone. If govts adopt it, they can use it as visitors list. I saw some QR code being used in a Singaporean university for classes. It might solve the proproblem of figuring out who attended a controlled social event.

On Tue, Mar 24, 2020, 20:25 Waqar Rashid waqarrashid33@gmail.com wrote:

May be we can convince small businesses owners to convince people to use the app, if they want to be informed about any possible infections from that shop/caffee. I saw lots of QR codes being used in china. We can look if we can find something about how they used it and if those can be applied here.

On Tue, Mar 24, 2020, 20:23 Waqar Rashid waqarrashid33@gmail.com wrote:

No. Only one phone would be needed for registering a contact. For example, a person going to a coffee shop can scan the coffee shop qr code. If other peoples who visited the coffee shop around that specific time and they turns out to be infected, we can have some potential infected people. I don't know if everyone would like to turn on their bluetooths. I personally keep my wifi and bluetooth off.

On Tue, Mar 24, 2020, 19:45 8lenka8 notifications@github.com wrote:

to scan users' codes from paper by third party?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/memiah/contact-tracer/issues/6#issuecomment-603439322, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABVDYAYH5ERJJJAKCZ724VDRJD5TZANCNFSM4LSORDQQ .

8lenka8 commented 4 years ago

But how do you track 'other people visiting that cafe'? you cannot take qr reading from some users with a smart phone and apply to others with basic phones who are near to them, because you can't track their location. They each would have to scan it with their smart phone or trace in BT or wifi - and permit us to submit it to our db...

wickywaka commented 4 years ago

Yeah, I get your point. We would have to rely on the decision of one user. Also I imagine that most of the infections are from interactions which lasts longer (don't know how much). I just want to make sure that what I have in mind can be implemented as an add-on. Lets see what others have to say on QR codes or if they can think of a better way. What I like about QR codes is that you can print it and then just have it like an ID.

There is an MIT project who are working on tracking locally and upload the data when the person is infected. That can help in backtracking an infected person manually. Not sure if include information about other phones.

On Tue, 24 Mar 2020 at 20:46, 8lenka8 notifications@github.com wrote:

But how do you track 'other people visiting that cafe'? you cannot take qr reading from some users with a smart phone and apply to others with basic phones who are near to them, because you can't track their location. They each would have to scan it with their smart phone or trace in BT or wifi - and permit us to submit it to our db...

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/memiah/contact-tracer/issues/6#issuecomment-603468422, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABVDYA77I343VAX6UPTV7EDRJEEZNANCNFSM4LSORDQQ .

-- Waqar Rashid

paulmaunders commented 4 years ago

Can you please fill out this form if you want to help? https://docs.google.com/forms/d/1ptWRKuay4UtlMiLWif1-cXUmokzJb00FDc2dnfxaTFU/edit

harshitsilly commented 4 years ago

@paulmaunders , @cmorillo - sorry to spam u both, I have already filled up the form and want you guys to help with the front end part

cmorillo commented 4 years ago

No worrries @harshitsilly , paul can you add him please I am making you admin on slack channel

harshitsilly commented 4 years ago

No worrries @harshitsilly , paul can you add him please I am making you admin on slack channel

still waiting @paulmaunders

harshitsilly commented 4 years ago

No worrries @harshitsilly , paul can you add him please I am making you admin on slack channel

still waiting @paulmaunders

still waiting @paulmaunders @cmorillo

harshitsilly commented 4 years ago

@cmorillo sent

cmorillo commented 4 years ago

Closing this as we will release a website soon and all inscriptions should be sent there.