Distributed Data Analytics / Edge Analytics

Position in the Platform
DDA is a core part of the overall FAR-EDGE factory automation concepts, which is classified as an enabler that supports the functionalities of the Analytics Domain of the FAR-EDGE Platform architecture. The specification and implementation of the DDA infrastructure take place in the scope of several tasks of the work plan, as illustrated in the following figure.

Tasks of the FAR-EDGE work plan that contribute to the realization of the DDA Infrastructure
The DDA infrastructure is directly related to the following constructs of the FAR-EDGE architecture:

  • The Open API for Analytics, which provides the means for accessing the capabilities of the Analytics Domain of the FAR-EDGE platform.
  • The Edge Analytic Engine (EAE), which enables FAR-EDGE to take advantage of the merits of edge computing such as the ability to collect and process streams in real-time and close to the field. EAE is itself a self-contained enabler, as it provides a runtime environment for executing analytics at the edge of any FAR-EDGE deployment.
  • The Data Publishing services, which are will provide the means for keeping the outcomes of the analytics (e.g., KPI indicators) up to date, through accessing timely data from the edge analytics functions (at edge gateways) in-line with their specifications. Each time there is a change in the analytics values and outcomes within an EG, the data publishing services will update the analytics outcomes in-line with the definition and configuration of the analytic functions.

Furthermore, the DDA infrastructure spans all tier of the FAR-EDGE architecture:

  • the Edge Tier that will provide the means for accessing and routing field data, which are implemented as part of the Edge Computing Infrastructure (ECI) of the project. Moreover, the Edge Tier will also host the EAE engine, which will provide the means for provides the means for executing locally scoped data analytics functionalities, and
  • the Ledger Tier is used to leverage “smart contracts” that will manage analytics configurations. A smart contract will typically keep track of information stemming from multiple Edge Gateway nodes. To this end, a proper set of Ledger Services are designed, in order to enable Secure State Sharing (SSS) across multiple EGs and their EAE engines. Ledger Services will, therefore, enable the implementation of advanced analytics scenarios, beyond locally scoped analytics at the Edge Tier.

High-Level Overview of DDA
A high-level overview of the DDA Infrastructure is provided in the Figure. It includes:

  • The Data Routing and Pre-processing component (including device registry service) have been implemented as part of deliverable D3.1.
  • The FAR-EDGE Distributed Ledger is used to orchestrate analytics functionalities across multiple EGs. Relevant functions that will be implemented as part of D3.4 (Secure State Sharing and Intelligent Reconfiguration) include Data Publishing Services and Configuration Services.
  • The Cloud Tier of any FAR-EDGE implementation stores data stemming from the DDA and provides opportunities for additional data analytics functionalities. At the Cloud Tier, one can also access the Open API for Analytics, which enables external systems to take advantage of the DDA infrastructure functionalities.
  • Edge Analytics Functionalities reside at the very core of the DDA module and run within the EG. They implement the EAE engine, which is the main contribution of the present deliverable. The EAE comprises the EA-Orchestrator, the EA-Processor and EA Storage modules, which are described in following paragraphs.


Distributed Data Analytics Architecture

February 12, 2018   1034    Articles    
Total 0 Votes:

Tell us how can we improve this post?

+ = Verify Human or Spambot ?