Implementing a next-gen repository mid-REF cycle…madness or a stroke of genius?

The University of Westminster has worked with Haplo over the past 4 years to build a comprehensive research information management system with functionality supporting researcher web profiles, research ethics and postgraduate research management. The Virtual Research Environment (VRE) has become a key research information system used across the University. For 3.5 years we have used a hybrid repository solution with a Haplo user interface and EPrints for discoverability. Benefits have included a high level of buy-in from our research community (they use the VRE for other research related activities), the improved user interface (alongside the REF OA Policy) has seen a huge increase in self-deposit, and streamlined workflows and improved open access reporting have presented opportunities to engage with researchers. If the hybrid solution was working...why take a risk so close to REF? The existing repository did not capture all of our research outputs (including practice based research and datasets) which made it difficult to showcase all of our research outputs or provide accurate data for reporting exercises. The one-way data feed between Haplo and EPrints is challenging to maintain and has made making changes difficult. Gaining approval to go ahead wasn’t a straightforward exercise. There was senior level trust in EPrints, concerns about REF reporting, security and the loss of the benefits of being part of the wider EPrints community. Why fix what isn’t broken? We needed to demonstrate how much better it could be, not settling for, “it’ll do”, that the risk of not changing was greater than going ahead. Not simply an implementation of an existing repository solution, it would involve working closely with Haplo to develop functionality, and there were time constraints - with a summer 2018 deadline due to REF timelines.