Loading...
 
Print

GEAF

GEAF is a project dedicated to build an autonomous and intelligent mission orchestrator prototype. The main goals of the prototype are reducing the human intervention in the data processing of PDGS missions and optimizing the quality of the processed data and performance of the instruments.

The GEAF concept is to be integrated with the MMFI Processing Facility Management and it will use as test case the three ESA Earth Explorer missions: SMOS, SWARM and forthcoming EarthCare.

The proposed solution for the GEAF system is shown below. It is composed by several modules that will provide the required functionalities in all the operation modes.

GEAF will have a dedicated and simple UI, to act as controller for the full GEAF system and to manage run configurations. Each of the GEAF Modules will report directly to the UI by means of logs.

The system will be able to work in three different possible configurations

  1. INTEGRATED: In this configuration the GEAF will control the MANAGEMENT LAYER in the MMFI. This means that it will need to detect relevant events in the PDGS, make decisions based on these events (issue warnings to the operators if required) and trigger the processors executions according to the orchestration rules configured.
 Image
  1. MONITORING: In this configuration the GEAF does not trigger any processor execution but shall work within the PDGS environment. This means that it will need to detect relevant events in the PDGS and make decisions based on these events (issue warnings to the operators if required).
Image
  1. STANDALONE: In this configuration the GEAF shall work outside the PDGS environment and will have all the capabilities of decision-making based on the orchestration rules. In this operating mode there is no need to have external interfaces, it is enough to have a pool of events that can be selected by the user in order to validate the behavior of the decision making engine. There shall be also a validation of the orchestration rules to identify potential conflicts between rules in either data-driven or time-driven constraints.
Image

Objectives

GEAF aims at demonstrating the maturity of technologies for the improved automation of PDGS end-to-end data processing .  The main objectives of the GEAF orchestrator are:

Analyse of end-to-end processing models for ESA missions: SMOS, SWARM and forthcoming EarthCARE in relation to processing facilities such as MMFI and/or G-POD.

Select most suitable technology and develop a prototype implementing the required automation.

Demonstrate the capability of the prototype to deal with end-to-end processing orchestration for selected missions.

Identify integration path for the prototype in G-POD and/or MMFI.

Application and Results Expected

GEAF will be validated with the current missions SMOS and SWARM and in case of successfull results it is intended to be used operationally by the forthcomming EarthCare mission.

The tool is designed to be generic and shall be easily configured to be used with other Earth Explorer Missions, allowing a better control of operations, improving the efficiency of the data procressing chains and minimizing the data contamination by orchestration faults. Another important benefit of using GEAF will be the considerable decrease of need for human intervention in the PDGS orchestration activities.


Contributors to this page: Michele Iapaolo

anasofiafreitas
.

Page last modified on Thursday 14 of July 2016 15:07:05 CEST by Michele Iapaolo.