Automation-Driven Quality Assurance strategy for research software as the vehicle to consolidate the European Open Science

• Autores: Pablo Orviz Fernández
• Directores de la Tesis: Isabel Campos Plasencia (dir. tes.) , Álvaro López García (codir. tes.)
• Lectura: En la Universidad de Cantabria ( España ) en 2020
• Idioma: inglés
• Títulos paralelos:
  • Automatización del análisis de calidad del software de investigación como vehículo para la consolidación de la Ciencia en Abierto en Europa
• Tribunal Calificador de la Tesis: María Teresa Ceballos Merino (presid.) , Germán Moltó Martínez (secret.) , Patrick Fuhrmann (voc.)
• Enlaces
  • Tesis en acceso abierto en:  TESEO  UCrea 
• Resumen

  • Given the strong endorsement by the research stakeholders worldwide, including national and international funding bodies and specialized agencies such as UNESCO, the Open Science movement is considered as the next quantum leap in science, decisive for accelerating innovation and knowledge sharing. In the particular case of the European Union, the transition towards Open Science has already started, undertaking the preliminary steps towards the establishment of an European Open Science Cloud.

    Open source software, in addition to open access and open data, has been identified as one of the main pillars of Open Science. Indeed, as modern research is requiring data-intensive computing, software is deemed as the cornerstone to drive the innovation in an increasing number of scientific disciplines. However, according to the analysis of Open Science-related literature conducted in this thesis, software is commonly regarded as a second class citizen, not being equally treated as the other research objects, such as data. This fact is particularly noticeable in the European case, subject of study of the present work, when examining the initial steps taken in the implementation of the European Open Science Cloud.

    This ecosystem aims at constituting an universal entry point for all the European researchers, offering them a catalogue of research-enabling services to do science according to the Open Science values. However, at the time of writing, these services are not being accurately assessed in terms of quality and maturity, and the role of the underlying software, that enables the services, has been disregarded to a great extent. As a result, there is an inadequate guidance and control of the minimum quality standards, which can pose the risk to the European Open Science Cloud of delivering unreliable services that not meet the researchers’ expectations.

    This thesis presents an automation-driven Software Quality Assurance process, built upon a DevOps culture, to govern the whole life-cycle of software production, from the early development stages to the final delivery as services, thus serving as a model for the quality assessment and integration of services in the European Open Science Cloud. This process is established on a strong foundation, evolving throughout the years of guiding the software delivery of several research e-Infrastructure development projects. The European e-Infrastructures have a solid background, forged over the years, on the development, operation and support of user-centric services, and accordingly, can be regarded as a benchmark for the delivery of services within the ecosystem of the European Open Science Cloud.

    As part of the outcomes of this thesis, concrete guidance is provided to drive the development life-cycle of quality research software. Furthermore, a quality assessment tool is presented, which is intended to contribute to address the aforementioned challenges with regards to the viability of the service onboarding in the European Open Science Cloud.

    Although the work is contextualized under the European scope, the outcomes and conclusions presented are equally applicable in a wider or global context. The main goal is to contribute to enhance the value of software in the scientific community as a key research object for the realization of the Open Science. Research software is the enabler of reproducibility in science, as well as an innovation accelerator in modern research. Only by ensuring the quality of research software, the aforementioned goals can be accomplished.