Use Mobile Data & Predictive Analytics To Prevent Spread of #Corona Virus

For most of us, our cell phones are our lifelines. Our cell phone witness and capture a lot of information about what we do, see, and share in our daily lives. We may not note or recall where all we have traveled and which route we took through the but a cell phone data can give accurate log of your activities to the second. , Cell phones are recording your activities every second and sending it to telephone company, app company and Google. As I have mentioned earlier as well, a small set of data by itself seldom has any value but by data analysis and correlation with other data sets we can tell a lot about people. There are these products called Predictive Analytics tools that can this data to predict your next move. Today every person carries a mobile phone and we use CDR (maintained by mobile company) & tower data (collected by mobile company for ech tower) to find how many people were around a particular cell phone user at any given time. (Read my older blog on How Google collects cell phone data including your voice data to know what we are dealing with today, but that's not related to today's post)

Let's assume a person 'A' has been infected with #Corona Virus  and another person 'B' was in close proximity of person A. Predictive analysis can scan through huge bulk of location data and highlight the phone numbers that were present in the locality at the given time. We can also trace their activity for a duration of time and create a profile of the person and predict

1) What is his residential area of the #Carona infected person

2) What are the addresses he frequented in last X days

3) What are the number of his close associates and by correlation who are their close associates?

4) What is a persons daily routine?

5) Does he own a personal vehicle or does he use public transport  (public transport has regular stops & fixed route)

6) By further correlation with data from financial institutions we can predict the persons credit-card or banking details

7) We can tell about his family members or the people who stay with the infected #Carona patient

8) We can identify the people who came within 1 meter of distance and could possibly be infected

Plan of Action for Govetrment of India

In the figure below if we analyze the CDR (Call Data Record of all mobile towers frequented by the #CoronaVirus patient and crete a virtual Geo fence of 1 meter radius around him, we can find how many people went close than 1 meter and so can possibly be infected by yhe  #CaronaVirus patient. All people who were in close prozimity of a infected  #CaronaVirus patient should be quantined as per the current goverment rule and kept under observation for 14 days. Mobile CDR can be analyzed using a Predictive Analytics software that can not onoly predict potential people who can get the virus but also show various trends of the #CoronaVirus. 

Suggested Plan:

1. Make a list of all confirmed #CaronaVirus positive cases
2. Get CDR data of patient from mobile companies ( Need Magistrate permission or blanket approval from PMO)
3. Get details of all mobile phone within 1 meter radius of #Carona Patient as ‘Potential Cases’
4. Ask local authority to contact all ‘Potential Cases’ and inquire about symptoms
5. Local authority & doctors should put all ‘Potential Cases’ on 24/7 monitoring for 14 days and check daily status 
6. Data engineer can write a software to analyze the CDR data and identify phone number of ‘Potential Cases’ who were in close contact of confirmed #CaronaVirus patient
7. Call Data Record  along with Tower Data record can be used to predict gorwth of #Corona cases and also help us in identifying ways to stop further infections 
8. Preditive Analysic has huge potential and can provide surprising predictions and there should be no delay in implementing Data Analytics
9. Government should take help of voulenteers from provate software companies to work on this initiative 

At the minimum a central  Data Analytics team of goverment of #India should be collecting data from all cities that have reported Carona Virus cases. The Data should be sent to the central team ASAP the city team gets some new data about the patient. This data set is cleansed and fed to the Predictive Analysis software to create a model and then predict various scenarios based on historical data ( in this case historical data can be just 3 weeks old) . The Call Data Record data can be collected from mobile serice providers like Vodafone, BSNL, Airtel & Jio on request from goverment.  private individual cannot get this data for security reasons. 

Is Function as a Service for your enterprise?

The name “serverless computing” means that the business or person that owns the software system does not have to provision servers or virtual machines for the back end code to run on. The provisioning of server is managed by service providers like AWS, Azure or Google and guarantees low latency content delivery network.

How does FaaS work?

Developers identifies a serverless architecture provider like Amazon Web Services (AWS) or Google Cloud to enable FaaS functionality for a web application. FaaS integration means some application code gets delivered from the 'Regional Edge Server' and availability and geographical distribution of Edge Servers is an important consideration. A user in India accessing a site that has FaaS code served from an server in Washington may encounter high latency that leads to high bounce rates. Taking example of AWS that has its global network divided as regions and within region edge locations that deliver service from the server closer to the web user delivering low latency and better user experience.So FaaS is very much closer to PaaS (Platform As A Service) , but there’s one difference. In FaaS, you don’t need to deploy the whole application in the cloud. Instead, you can deploy a part of it as a service.

Lets take example of few use cases for various vendors and their serverless offerings

Amazon Lambda

Lambda is a serverless computing offering from Amazon Web Services (AWS). Many of you will wonder what this is. Let me tell you that this offering from AWS will change the way application development and its cloud hosting is done.With Amazon Lambda, you can directly run your code without even managing the servers. For example, you upload an image in the server, now the Lambda function will work for you and will automatically resize the image according to the device a user is using, be it mobile, laptop, desktop or a tablet.To use Lambda services, you need to take a plan that will be “pay as you use.” This means you’ll need to pay only for what you are using.AWS Lambda also helps developers to scale the applications by running the required codes whenever someone triggers.

How does serverless computing works:

1. Developer writes modular code for a functionality in the code editor.
2. Developer deploy the code & configure it to be triggered by other AWS services or in-app  activities/events.
3. Once code is deployed AWS Lambda will run the code when it gets triggered by an event or another service.
4. On AWS Lambada you only pay for the compute time of the deployed code.

Google Cloud Functions

Google Cloud Functions is a serverless computing service launched by Google in 2018. Google cloud functions enable developers to write codes to extend and connect to cloud services i.e., Google’s public cloud. With Google cloud functions, website developers can call particular functions which can perform specific tasks which when triggered gives specific response. These responses are automatically triggered by the Google Cloud Platform (GCP). To provide a seamless experience to the developers, Google introduced the service Google Cloud Functions & made it easier for the developers to to  focus more on the coding part rather than developing infrastructure for the application. Scalability is also one of the significant throwbacks of this service. This service automatically scales the resources whenever they are needed and puts them aside when not. With the presence of conventional coding languages like JavaScript and Python, it is easier for the developers to develop the codes and upload them according to the functions.

Microsoft Azure Functions

Azure Functions is the FaaS launched by Microsoft. Microsoft Azure Functions are designed by keeping the developers in mind. This service helps the developers to accelerate the application designing and development process. Microsoft’s idea behind this FaaS is to eliminate the time consumed by the application infrastructure development process. Now when this is eliminated, the user can easily create the software application and upload its code.  The users now have to pay only for the time which is consumed by the function to run. Also, this service is hosted on Microsoft’s Public Cloud.

IBM Cloud Functions

Based on Apache OpenWhisk, IBM Cloud Functions is a polyglot functions-as-a-service (FaaS) programming platform for developing lightweight code that executes on demand. FaaS helps in developing lightweight codes that are executed when demanded. The service is very similar to that of AWS Lambda and Google Cloud Functions.

Some other popular examples of FaaS are Iron.io & Webtask.io

Summarizing

FaaS lets developers write and update a piece of code on the fly, which can then be executed in response to an event, such as a user clicking on an element in a web page. This makes it easy to scale code and is a cost-efficient way to implement microservices.

Advantages of using FaaS?

Improved developer velocity
With FaaS, developers can spend more time writing application logic and less time worrying about servers and deploys. This typically means a much faster development turnaround.

Built-in scalabilitySince FaaS code is inherently scalable, developers don’t have to worry about creating contingencies for high traffic or heavy use. The serverless provider will handle all of the scaling concerns.

Cost efficiencyUnlike traditional cloud providers, serverless FaaS providers do not charge their clients for idle computation time. Because of this, clients only pay for as much computation time as they use, and do not need to waste money over-provisioning cloud resources.

Drawbacks of FaaS?

Less system controlWhen a third party manage part of the infrastructure it is tough to understand the whole system and debigging becomes more challenging.
Testing becomes more complex  as it is  very difficult to incorporate FaaS code into a local testing environment, making thorough testing of an application a more complex task.

Monitoring become difficult as 

The idea of serverless architectures is appealing or developers. Having sai that FaaS cannot be a complete replacement for traditional application architectures.
To build a basic web application as FaaS, it would take a lot of functions. Application that are using microservices architecture are ideal for replacing potential microservice style architectures and background type services.

Experience Level Agreements (XLAs)