Prototyping of a machine learning based test data processor for autonomous testing processes from test design to report evaluation, applied to ground and space systems.

While test case definition is test dependent and test performance is human error prone and effort intensive, automated testing proved to be still not reaching a level of abstraction and exhaustiveness to enable significant return on investment.The application of Data Analytics technologies like Machine Learning, Artificial Intelligence or Data Mining to both simulation and test data processing shall provide a level of abstraction and consistency to improve both test automation production and exhaustiveness. The would lead to close to zero effort to generate an automated test case from user inputs that can be run at discretion.It is aimed at inserting Data Science to:1. test case design: definition and recognition of patterns from user inputs2. test case automation: organisation of patterns into series of nano, micro and macro building blocks3. test case performance:integration of test cases into existing test frameworks4. test case assessment: definition and evaluation of pass/fail test criteria, in particular spurious and sporadic behavior recognition5. test case report: provision of test results including source cause identification and resolution guidelines based on previous test reports and problem resolution reports into existing frameworksThis activity includes the tasks:- Analyse test processes, workflows and users needs to identify capabilities and requirements- Identify and evaluate potentials for optimisation among test executions and needs- Assess state-of-the-art technologies to support automated test processes across domains- Identify opportunities to optimize test execution- Identify representative pilot cases to apply the prototype in a representative environment- Develop and implement the prototype to demonstrate the pilot cases- Evaluate benefits based on criteria to be agreed - Identify areas for follow-upsExisting candidate framework to be upgraded with these technologies are:- Building blocks for End-to-End VALidation and management of distributed ground segment systems components (E2EVAL - G617-231GI)- Test Report Standard and Advanced Test Management System (ATMS - T708-505SW) activity results.The resulting Prototype Software shall be delivered under an ESA Software Community Licence, so that any legal entities within ESA Member States can access, upgrade and contribute to it.