Building a virtual research environment (almost) from scratch for the Heritage Science Data Service

Heritage scientists frequently need large-scale datasets and specialist software but often lack the compute and access to it. Many heritage scientists have graphics intensive workflows, and 3D models too large to use on their laptops. RSEs bridge this gap in many fields, and that is why we have built a Virtual Research Environment (VRE) for the Heritage Science Data Service (HSDS). HSDS is a partnership between the Archaeology Data Service and STFC Hartree Centre, with a consortium of key UK heritage institutions, delivering core digital research services as part of the £80m RICHeS programme. Our VRE provides browser-based access to accelerated compute, integrated datasets and software, with ease of access being the core requirement. Unlike a Trusted Research Environment (TRE), we must make it as easy as possible to bring data into and out of the environment. As ever, our challenge has been translating users’ requirements into technical decisions, stitching the right open-source components together while building our own. This talk will walk through a few key decisions where user needs have shaped our architecture, from streaming virtual desktops to staging large datasets in and out of object storage. We will show another aspect of an RSE, building bespoke environments where research takes place. Whilst it’s not every day that you build a VRE, we will share some lessons learned that we think are transferable to many RSEs.