Cross-disciplinary and cross-life-cycle-phase Technical Debt in automated Production Systems: two industrial case studies and a survey

Technical Debt (TD) metaphor was introduced at classical software engineering domain more than two decades ago. Outcomes of TD research have been gradually showing the benefits, which support practitioners to identify and remove TD items, thus reduce the maintenance cost. Recently, TD concept has just been successfully introduced into automated Production Systems (aPS) domain. aPS have long operation time, which has significant maintenance cost. In aPS domain, the research about TD is more complex, because besides software, aPS consist of other disciplines such as mechanical and electrical parts. Strong relationships of these three disciplines have substantial influences to development and maintenance phases of aPS. Thus, it is a challenge to study TD items in this domain. We need a broader investigation with involvements from multiple departments, to examine TD in aPS, rather than just analyze software part. However, up to now, most researches about TD in this domain still focus on software part only. This research studies TD items in aPS concerning the surround factors and impacts of these factors to the software side. The research method is TD4aPS, an adaption of SWMAT4aPS benchmark process to TD research for aPS. We select and present two typical case studies in the paper. One case is from a machine manufacturer. Another case is from a plant manufacturer. Two typical characteristics of TD in aPS domain are shown: cross-disciplinary and cross-life-cycle-phase. From analysis, the detected TD items are classified then short- and long-term recommendations are proposed. We validated qualitative results in case studies by quantitative results from an online survey received from German industrial companies working with aPS at 40 different markets.