Abstract :
[en] Over the last decades, research has become more digital, collaborative and open. Research openness has extended from articles to other research outputs, including datasets. Meanwhile, datasets are growing bigger and are requiring more financial and energy investments. In addition, the reproducibility crisis has highlighted the need for a cultural change in research data management, which most funding agencies are nudging by requiring compliance to the FAIR data principles. Institutions can help researchers address this new pressure through support from cross-disciplinary staff and librarians, and by implementing general data management tools. This basic and necessary initiative from universities most often translates into awareness campaigns and general training sessions that often end up overbooked, highlighting the growing interest of researchers in these issues. However, this central approach lacks the discipline-specific expertise needed to properly translate the general recommendations into actionable items. With most institutions being decentralised across many campuses, logic pushes towards relying on local relays. Inspired by their peers from Cambridge and TUDelft, the six universities in the Brussels-Wallonia Federation (FWB, Belgium) have launched, as a consortium, a community of Data Ambassadors (DAs). Sometimes called Data Champions, DAs are researchers acting as local experts who bring awareness in their immediate work environment (department, research unit…) towards data management best practices. DA networks have the advantage of automatically addressing another weakness of central support, that is, the lack of resources, by capitalizing on existing workforce with the relevant expertise. While lean in terms of spending, this approach does however point towards the issues of resources availability for such services. On top of these benefits, DA networks enable peer-to-peer support, which has been shown to be a much more efficient drive towards change than top-down initiatives. The FWB consortium hopes to strengthen links between research groups across universities and disciplines. The aim is to empower individual researchers and engage whole communities to be better data managers, instead of trying to command change from a hierarchic point of view. Launched in December 2021, the DA community has already received massive interest with over 60 members enrolled over the first couple of weeks. Although inspired by other institutions, this network presents the specificity of being an entirely bottom-up initiative, without any other existing catalysing interuniversity structure, thus facing the challenge of building its own tools from the ground up. The following paper introduces the launching process - based upon the TUDelft model, and it reports on the overall experience from pioneering DAs, including their achievements and difficulties. The vision for settling and expanding the network and its expected successes and challenges are presented.
