Abstract-Knowledge-based systems, are today part of many research protocols where they act as powerful means to model, implement and cross-examine the workflows that lead from a set of inputs to a set of outputs.They remain however tricky to apply in the specific context of heritage science where workflows include a long tail of subjective human decisions, of non-explicit research protocols, of poorly formalised pieces of knowledge, of highly individual skills, of undocumented, non-reproducible, intuitive interpretations, when not simply of licentia artistica. Yet the heritage science community has witnessed over the past decades the emergence of huge quantities of digital outputs, either following massive digitization efforts, or as a result of the growing capacity of actors to produce digital-born material. How can this move be supported in terms of reproducibility, reusability and cross-examination of results if research protocols remain non-formalised one-shot efforts?The research presented in this paper bases on the idea that what should be formalised and shared with future generations are not end results alone (outputs) but the methods and processes that lead the making of the output (human skills, tools, technological procedures, cognitive processes, scientific protocols, etc.).Our contribution addresses a pending issue: how can we today complement traditional approaches to heritage assets documentation with means to describe and record research processes and workflows? The infrastructure we propose raises knowledge representation, visualisation, and information management issues. It applies primarily to a range of specific cultural heritage related artefacts, but is expected to be fairly generic in terms of methodology. In this paper we describe the methods employed in order to elicit underlying activities, support team elicitation through ad-hoc visualisations, promote a consistent visual interfacing of the underlying Information System.