Atlas Simulator - A Pandemic RealTime Simulation Platform - with options to amend all types of "Live data"

Short Pitch

The World needs a pandemic/epidemic "simulation platform” that is realistically modeled by the worlds population and their habits - individually and as a herd. As long as we don’t understand how a virus spreads in the population, we must continue to work with policies and action plans that are inherently uncertain, such as closing all areas for public mobility when there can be significant drops in risk that must be weighed in. We can become more exact in our risk assessments by applying real-time simulations. Furthermore, the simulation itself, and the data it outputs can be used in other applications to make assessments of the risks of going somewhere. The wider benefit of integrating the simulation with other applications becomes how it is distributed, to give insights to a wider audience in a society struck hard by an outbreak. This can happen via apps like Google maps, to give the user data on what stores, grocery markets, have had a bigger influx of people and hence provides a greater risk of going there.  The simulation could of course also be run on a standalone app, available for anyone to download.  The key takeaway here is that the simulation takes in real time data so these feedback reports would not lag by weeks or days, but it would be continuously updated on a minute to minute/hour basis.

Problem Description

The reason for the mass shutdown of our production output, local stores and enterprises of scale, lies in the notion that the virus could be anywhere and so it becomes very hard to make a risk assessment of what can be open for the regions where risk is the lowest, hence the economy tanks. Furthermore, people feel isolated and afraid since they do not know where they can safely be, how present the virus is at the location they are present in. There is no real world data updating our citizens by the minute or even hour. A strong feedback report is therefore essential but today non existent for people to use that is based on data that’s also unique to their situation. We need a more enhanced way of tracking the spread of the virus and identifying where the risk of spread is lower. It’s not like we have a simulation platform adaptable for the demographics data of each city, each region to determine more precisely how the virus spreads.

Solution Description

By creating a simulation platform for how a virus is spreading throughout a region, we can not only get a better understanding for how some parts get more affected than others but we could also create a situation where enough data is gathered to confidently say that this region has such a low virus spread, and such a low risk of creating additional cases, that it can stay open for business. The obvious effects of this is more stimulus to the economy. To have a policy of “let’s shut down everything” lies in the fact that we do not have enough data to know exactly how the risk fluctuates, region by region. Furthermore, integrating the simulation into high usage applications can provide big rewards for society as a whole. For example, no matter if we do a standalone app or integrate the simulation to run in high usage apps like Google maps -  if someone searches for nearby grocery stores then the simulation can do one episode in the background, and from that episode give feedback that although this store is the closest one from your current location, there has been more foot traffic there which have made that area more susceptible for virus spread. More features can also be added to an app. With a simulation so exact with its virus spread assessment that it can give a lower risk rating for an area within a region with a much higher risk rating, a warning in the form of a notification or even a beep can be sent out via the speakers on the cellphone via the app, when the user is approaching a region where the risk rating has changed, i.e if the user is going from a lower risk area to a higher risk area. This is just one of many use cases for a complex simulation that is being fed real time data.

One crucial part of creating a sustainable simulation platform is to have strict privacy laws. The data we get on the location of people will be strictly geographical location data.  We care about herd movement, not individual cases, person to person. Technically this will be done by encrypting the incoming location data of people, chunk it up in pieces to ensure that we minimize the risk with whatever means we have from any raw data to leak out. Furthermore, GDPR will be taken seriously and followed in any possible way we can.  The data we take in will be properly stored and only accessible for a clear cut time period before it gets permanently deleted from all of our systems. No data that is not relevant for our core business will be stored and the data that is stored will follow strict protocols for how we handle that data. Overall, this project and all its actors, strongly support data privacy legislation and strong privacy practices will only strengthen our platform and increase its reliability. Moreover, we are all educated to a sufficient degree in GDPR to know that we must all respect it and to know how we should handle our data. How we collect, store, and transmit data will always follow strict GDPR guidance protocols. This applies to all current, and all future operations. Other ways we would like to comply with GDPR in the near future is to hire a Data Protection Officer to oversee our data practices and offer advice on how we can improve.

Step 1 - Year 2020→2022) To have an Open Sourced, Freedom for everyone to use and contribute to in simulating pandemic influence on our planet in RealTime … so we can make better informed decisions with the ability to “measure” the “outcome. Thinking of Anders Tegnell here, when his team have “data” lagging several weeks behind, having a “hard” time understanding what effect an action had.

Step 2 - Year 2020→2025) Able to do “Contact Tracing” of infected people and having that “data” live rendered in the “Platform”. Where people affected by a virus could get a notification on their phone.

We don’t need “Yet, another App for that”! It is instead crucial to work with Apple & Google to implement the “notification” to End Users in their Operating System. It is better said information comes from a credible source, from the suppler they have chosen to purchase the device from.

Step 3 - Year 2021→2030) This is next step-stone is a “Proof of concept” idea … and stems from having  worked with IOT devices, mainly on “Asset tracker” for +2-3 years.

The “missing” piece of the puzzle is a “digital virus sensor”. KTH [1] are working on a solution, though we have some creative ideas of our own on how to make one.

When a viable “digital virus sensor” is available, we could install them in all of the  “Public Transportation” vehicle people use, every day. When a persons sneezes, and their “exhaled air” is moved around, and a “droplet” with a virus lands on the sensor … the device detects it and an alarm is triggered and send directly into the Atlas PSP, and proper action(s) can be put in place.

We see a future, where we’ll be able to “extinguish the fire” before it ever becomes an Epidemic or a Pandemic.

Action Plan

We have developed an approach that takes into account the demographics data for each city. More specifically - data on the population, urban settings, and perhaps most importantly; the foot traffic of a region, i.e how a population moves within a region. With this input we can adapt the simulation for each city that have demographics data. Thereafter we can use the data to simulate where the high risk hotspot locations are. When we know the hotspots we can add a “hotspot” parameter to the simulation to simulate probability of spreading the virus to other parts of a region, depending on what quantity of people visited a hotspot and for what period of time. With the findings that are generated from the simulation users can realize in a visual way how they act as a mechanism of spreading the virus while getting reports based on real time data on how the location they are currently in have been exposed to the virus. The simulation will also provide insights for its users on why a certain area now has a higher risk rating for virus susceptibility and what places can safely be visited instead. The benefit with data output of this magnitude is that people will not have to guess if a place is safe to go to. Based on more exact findings of virus spread from the simulation data, a person will know via the feedback coming from the simulation if a local area, within a region more heavily affected, has a safe rating or not. The consequence of output like this will then be data that can support more public mobility, but, in a safe manner which will decrease isolationism for a large group of citizens.

  1. Gather demographics data (with a focus on foot traffic)
  2. Gather data for how we normally move around in our society. When we don't have an epidemic.
  3. “train” simulation runs on data gathered
  4. Run the simulation in real time
  5. Look at the pace of spread for a city and extract probablity for another region to have a similar spread
  6. Isolate cases with a high probability of spreading the virus
  7. Use CFD*, virus backtracking among people movement, and add techniques to the platform for enhancement
  8. (Use simulation to enable an app to tell you what areas are high risk and what areas are low risk)
  9. Also integrate simulation platform with mass use applications like Google maps for assessments on being close to areas, or going to places where a lot of people have been that develops a scenario of higher risk, in relation to a virus outbreak

*It is nowadays common practice that we use “Computational Fluid Dynamics”, CFD, software to simulate and test cars, aeroplanes, busses, trains, bridges, motor cycles, formula 1 cards, how to clean and fill folding carton packages our food is packaged in (Having work experience  at TetraPak, the Tetra Evero was the 1st time they began using CFD for this purpose) and the list of what CFD’s are used for is endlessly long.

Demo Video

https://bit.ly/Atlas-Sim_PresNDemo_002

Tech Description

We use python backend and mami API to get visuals on the actual simulation. CFD applied for effects of sneezing, coughing in a confined area with people. Perfect use case for mass transit for example.

Please note, that this is “Right Now” - and we plan to port the code to C/C++/Swift, and make it available on all types of operating systems and devices. We’re going to have an “agnostic” operating systems approach … “It should run on anytime (TM)”

Github Repo

https://bit.ly/Atlas_Sim_Git

I understand that this plattform will be running from Github:

In this doc there is a need of clarification of the owners of the plattform;

License

Free and Open Sourced under the MIT License, granted by Grant Sanderson [2], when forking his codebase: “manim - Animation engine for explanatory math videos” [3]

Ownership

Like Linus Torvalds has done with the “Linux kernel”, so shall we do with this project.

Daniel Eftodi of Absolute Creativity Design Studio, with the Team that he has managed to attract & assemble during the #Hack-the-crisis-Sweden Hackaton, will oversee that the projects goals will be upheld and fulfilled, as to their best ability and time.

PR

Please contact Daniel Eftodi

Stakeholders

  • Public Health Agency of Sweden (Folkhälsomyndigheten), FHM
  • National Board of Health and Welfare (Socialstyrelsen)
  • Swedish Civil Contingencies Agency (Myndigheten för samhällsskydd och beredskap), MSB
  • All regions of Sweden, and any other Country
  • The Swedish Parliament (Riksdagen)
  • All Governments of planet Earth
  • All Companies of planet Earth
  • Every life-sign of planet Earth

Needed Resources

Skills: Anyone willing and interested, Hackers, among them key-stakeholders with prior knowledge (eg FHM, MSB, Socialstyrelsen, Scientist, Doctors, etc etc etc)

Competencies: Anyone willing and interested, Hackers, key-stakeholders with prior knowledge (eg FHM, MSB, Socialstyrelsen, Scientist, Doctors, etc etc etc)

Investments / Finance: From interested parties and stakeholders whom would like to see this through to fruition

Functions;  for example Steering-group, team, controller
Equipment / Online:

Team Description

We are serendipity found each other when Daniel Eftodi posted the “idea pitch” in the slack-channel. From then on, we began working together on this common idea we all share and believe in. One-by-one our Team expanded to around seven:ish people, and have found a great Mentor that shares our passion in brining this idea to fruition - YAY-TEAM

Team Contact Info

Daniel Eftodi, Inventor & Product Developer

Stellan Lange, Project Developer AI

Joakim Pettersson, Systems engineering, innovation, programming, testing

Nagaraju Erigi, Data Engineer

Johan Thor, Senior Medical Advisor, Associate Professor

Fernanda Dorea, Veterinary Epidemiologist

Leyla Avzel, Our Mentor

References

1 - https://www.kth.se/mst/research/sensors/project/detection-airborne-1.523039

2 - https://www.3blue1brown.com

3 - https://github.com/3b1b/manim

Try It out

Hackathons

Technologies

c, c++, grit, livedata, python, swift

Devpost Software Identifier

259346