Similarity-Based Process and Set-Up Time Estimation

Contract manufacturing companies have to calculate offer prices for individualized products in cost-driven markets with experience-based estimations on limited data. Especially, the estimation of processing and set-up times is time-consuming and challenging. Algorithms providing suitable and transparent estimations with low configuration effort and licence costs for several machining processes are currently not available. Therefore, a method is developed to provide automated estimates for machining and set-up times from production data based on similarity metrics. This method is evaluated using feedback data from turned, milled and whirled workpieces. Experiments show that, compared to experience-based methods, an accuracy increase of 7.8% is achievable for process time estimation of turned parts and 21.2% for set-up operations.