The digital transformation is omnipresent in science. Especially engineering oriented scientific disciplines have been used to computers as a valuable tool since they became available. The wide availability of computers and computer applications in the end of the last century lead to a fragmented development of non-interoperable computational solutions. In engineering oriented scientific application those solutions have typically targeted the answer of a single well determined problem. Recently the technological evolution provided the means of generating added value by connecting data and drawing new information from it. But not only data, also people and their know-how has to be connected. We describe how this challenge is tackled in our typical long tail data scientific organization in the domain of material science and engineering. To implement this transformation process we established a crossfunctional unit, the HUB MaterialDigital. Permanent members of the HUB are scientists and consultants, working part-time in their home-divisions and part-time in the HUB. Scientists from the operational units, the domain experts, join the HUB with specific domain problems for a limited period to implement their special solutions. The role of the HUB during this cooperation is to accompany the creation of ontologies and other standards, to master the tools and platforms and to control the implementation process into the institute’s digital infrastructure. The role of the domain expert is to assure correctness in terms of the domain and to prioritize the domain specific functionality by formulating and driving requirements. All members share an open plan office during the project. The team has adopted a Scrum-process to handle projects: there are core times where sprints, short update meetings, are held. This enables the team to act fast and flexible according to actual development-challenges and especially to the domain specialists’ needs. The first project experiences show that the adopted approach led to an efficient handling of digitalization projects. The expectations towards the results quality and project duration were met: the projects were finished significantly faster than comparable projects that use the traditional isolated workpackage approach.
Rebecca Reichenbach (Fraunhofer Institute for Mechanics of Material IWM)