Use case - dataflows

Get started

Quick up-and-running

In Get started you can find a description to get quick up-and-running on a Linux machine:

• Prerequisites
  • A machine with Ubuntu and how to install Docker
• Tidepool
  • Create a local instans of the platform with a local Kubernetes cluster
  • Install uploader and web
• Semantic Container
  • Install semantic container

Integration

Extract data from device using Tidepool and use in Semantic Container dataflows

Setup a single Linux integration demo
When finished uploading then you can do the following: (refer to the orange number in figure):

1. Grab the data for further use in the dataflows
2. See the data in Tidepool Web

In the dataflow Tidepool integration you can find a demonstration of these capabilities:

• reading data from a Tidepool supported device
• avoding duplicates on data upload into a Semantic Container
• providing a Diabetes specific Usage Policy
• generating a Provenance trail specific to diabetes devices

Personal data

Store data to Semantic Container and get insigths in PDS

In the dataflow Personal_Data you can find a demonstration of these capabilities:

• converting diabetes data into FHIR-compliant RDF format
• accessing data through a SPARQL endpoint
• uploading data to the OwnYourData Data Vault
• strong encryption of diabetes data
• authentication methods for diabetes data
• visualizing data in a Knowledge Graph
• gain insights from personal data through individual analysis
• access and process diabetes data with standard tools

Data donation

Share data and identify if data leak

In the dataflow Data_Donation you can find a demonstration of these capabilities:

• avoding duplicates on data upload
• providing a Diabetes specific Usage Policy
• generating a Provenance trail specific to diabetes devices
• matching Usage Policies specifically based on subsets for Data Categories
• applying and verifying digital watermarks

Data tracing

Share data and aggreate/anonymize data

In the dataflow Data_Tracing you can find a demonstration of these capabilities:

• PwDs upload their local diabetes data to an aggregation/anonymization service
• aggregating/anonymizing data in a Semantic Container
• a service provider transfers selected (aggregated) data to a 3rd party
• each PwD traces his/her own data and compiles a usage report

Proxy

Use a proxy to handle integration

We have defined a concept of a health data store (HDS) to merit of higher protection for sensitive data as health data. For scenarios of multiple users in the same environment we have developed a template that can be used to demonstrate the workflow as swagger response.
In Proxy you can find a demonstration of these capabilities:

• template to build a demo with Python using Flask web application framework
• local demo and upload to a docker repository available for the kubernetes cluster
• design and security considerations for deployment for building an online demo