Loading...
 
ESA > Join & Share > loose-methodology
Print

loose-methodology

The LOOSE project develops and evaluates technologies and solutions for interoperable processing and timeseries analysis in online platforms derived from contemporary use cases and stakeholder needs.

The project work is organized in three tasks:

  1. collecting requirements and defining scenarios from identified stakeholders.
  2. refine and implement the architecture in two iterations (cycles).
  3. evaluate the developed prototype in collaboration with the project stakeholders.

Therefore the tasks are broken down into workpackages that are connected by a main project workflow as depicted in the above figure. The workflow is a sequence of the following activities:

  • Scenario Definition: Analyse stakeholder user stories and needs; derived common Use Cases; define EO exploitation scenarios that shall be implemented within the LOOSE prototype and evaluated with the stakeholders
  • State-of-the-Art Review and Analysis: Investigate promising contemporary technologies, standards and operational online platforms for discovery, access, processing and analysis of long EO time series; identify limitations in existing tools and standards
  • Requirements Definition: Derive user and system requirements based on the technology assessment and stakeholder scenarios
  • Architecture and Design: Define a system architecture for the LOOSE Prototype with respect to the requirements
  • Prototype Implementation: Develop interoperable, scalable solutions for large-scale systematic processing and on-demand analysis of EO datasets and products; enhance Open Source Software and tools to overcome limitations in existing implementations;
  • Integration: Continuously integrate the prototype components in the Reference Infrastructure and verify the functional and non-functional requirements as version 1.0
  • Deployment: Distribute and install the prototype components in the Prototype Infrastructure on a selected online platform as version 2.0
  • Evaluation: Validate and demonstrate the prototype based on real-world scenarios in selected operational environments with the project stakeholder
  • Outreach: Actively participate and collaborate on standards in standardization bodies, working groups and community activities; demonstrate prototype solution on relevant eventsand collect feedback for refinement of the prototype

The Stakeholders are actively involved in the definition of the use cases and scenarios and during evaluation of the prototype in the Prototype Infrastructure. However it is also intended to collect valuable feedback regularly during integration and verification of the components in the Reference Infrastructure.

The Online Platforms capabilities and technologies are researched in detail during the State-of-the-Art Review and Analysis. This is important to understand the platform differences especially regarding the interfaces for requesting and billing the resources as well as deploying the prototype. The review results in system requirements for the architecture and design as well as a selection for the target validation environment for the prototype (Prototype Infrastructure). Feedback on the performance, cost and usability of the platform is collected and reported in the evaluation activities.

The Standard Organisations are closely consulted and involved during the whole project. In Task 1 relevant standards such as WCS, WPS and OpenSearch are reviewed and their current state and potential evolution is discussed. In this context also potential harmonization with alternative initiatives such as the Spatial Temporal Asset Catalog (STAC) and OpenEO are evaluated.

The prototype is used throughout the project for demonstration purposes, outreach activities and interoperability initiatives. The prototype will be developed in multiple development cycles, ultimately resulting in two versions: v1.0 and v2.0 (see Figure 2). The main goal the v1.0 cycles are to integrate all individual components in the project-internal Reference Infrastructure to verify the developed software (extensions) and ensure functional compliance to the defined scenarios.

However, in contrast to v2.0, no formal validation is carried out. Instead the agile development in this phase guarantees rapid feedback to the developer from scenario integration tests as well as demonstrations to stakeholder and standardization organizations. While this version will result in a full-stack prototype, the scalability and performance is restricted due to the underlying infrastructure and reduced availability of datasets.

Build upon v1.0, the development cycles resulting in the final v2.0 focus on cloud deployment, scalability optimization and development of minor extensions. This version is deployed on a public cloud infrastructure (Prototype Infrastructure) supporting large-scale processing based on long EO timeseries. Therefore this version is the foundation for detailed validation, usability and benchmark tests executed as part of the in-depth evaluation with the project stakeholders.


Contributors to this page: Bernhard Buckl
meyerarnek .

Page last modified on Tuesday 07 of July 2020 10:00:33 CEST by Bernhard Buckl.