Scrum within a scientific development environment

Commercial successful software products that address a specific scientific domain usually started out as an exploratory prototype written by group of subject matter experts with strong scientific background who have acquired the programming skills, transforming their theoretical formulas into working code. Research and planning typically do not marry well. Particularly, the traditional waterfall approach with its strong focus on a predefined requirements scope, design and implementation phase, does not appeal to research staff who find it difficult to predict what they are going to build.

This research showed that the lean ASDM approach seems to be a better fit for research developers working on scientific software solutions for it embraces change. It was obvious that the iterative and incremental lifecycle approach that Agile supports, proved to be much better tailored to the research way of development in that it allowed for timely changing direction if an idea does not yield the expected result or users provide different requirements. The same benefits applied when using research stories which are often difficult to manage.