In steel fabrication and erection, the production flow is measured by counting the amount of fabricated and erected tons. This leads to several questions: is this indicator sufficient to visualize the real productivity? Are there better metrics to display process performance? Thirdly, how can lean methods like pull, Kanban, ConWIP and SMED improve these processes? A study of steel fabricators in Germany and Chile revealed that the only KPI today is the global metric of [hours/ton] related to the finished products. The lack of performance metrics, high process variability and individual production justifies the first research question concerning whether there are common patterns behind all steel-fabrication processes. Further questions involve finding the right metrics and determining which lean-methods could be best applied at certain work-stations to optimize the whole production flow. Data is measured in detail considering the whole steel-fabrication process including the erection of the building on site. This paper provides an exemplary in-depth exploration of the cycle times and process variability of the work-stations conducting ―fitting-welding‖ operations. The application of PULL through Kanban and CONWIP as well as takt-time control and SMED is discussed. A three level KPI-concept is elaborated, to support pull in the whole supply chain. Finally the study provides a basis for simulating the steel fabrication and erection process as tonnage flows by continuous simulation (HECRAS) or the one-piece concept of DES-Simulation (SPS).
Steel-fabrication process, Production metrics, Process variability, Standardisation, Product complexity, Continuous- and discrete event simulation