Experiences from Bottom-Up Python Software Development in Experimental Physics

Abstract

Automation has become an important aspect of effective and rigorous experimental research work, particularly in physics. Among the diverse automation platforms, Python stands out with its large repository of open-source scientific software for instrument control and data analysis. However, despite considerable effort from the research software community to unify multi-instrument control (also known as instrument orchestration) and to make it laboratory-agnostic, there has yet to be a strong consensus on a universal package that is easily adaptable. We contextualise this issue by discussing the underlying barriers that we have encountered within the experimental physics community, such as the unfavourable circumstances for software development, and the disparity in programming skills amongst physicists. A sustainable way forward could be specialised but well-maintained software repositories within each research group. In that spirit, we present our experiences as case studies on building software for experiments and share key coding considerations that may be helpful to other physicists.