Experiences with technical debt in parallel multi-disciplinary production systems engineering

Abstract

Technical debt (TD) is a well-researched area in software engineering (SE), to communicate the long term negative effects and impact of short term solutions. Most TD research focuses mainly on software code and its implications, whereas in other disciplines, such as parallel multi-disciplinary systems engineering, the engineering of large and long-running automation systems, TD is a relatively recent concept. A central similarity of production system engineering (PSE) to SE is the integration of partial results from different workgroups, who are working simultaneously on data artifacts which need to be combined later in the process. To better understand this business process and in further steps improve it, we analysed TD in the context of engineering data exchange. We identified two main categories Technical Debt in Knowledge Representation and in Exchange Process. Re-enforcing relationships were identified, similar to contagious debt in software engineering, that reproduced to other workgroups or later phases in the process, as well as multiple causes for TD. The research approach was a case study in the context of exchanging engineering artefacts in one PSE company to identify TD causes, items and effects through technical debt analysis. We validated the results a) in a qualitative way by conducting semi-structured interviews, with representative from the investigated company and a similar company to refine and test findings and b) through quantitative results from an online survey with practitioners. The initial results, although of small sample size, showed promising results in the industry domain and for multi-disciplinary industrial contexts. The results indicate, that our TD taxonomy is applicable for other companies as well. More validation and focus on other domains of TD could be topics for future research.