With regard to the recommissioning of damage caused inoperable complex capital goods, a high logistics efficiency is a very important competitive factor for regeneration service providers. Consequently, fast processing as well as a high schedule reliability need to be realized. However, since the required regeneration effort for future damages may vary and is usually indefinite at the time of planning, capacity planning for the regeneration of complex capital goods has to deal with a high degree of uncertainty. Regarding this challenge, the evaluation of prior regeneration process data by means of data mining offers great potential for the determination of load forecasts. This paper depicts the development of a data mining approach to support capacity planning for the regeneration for complex capital goods focusing on rail vehicle transformers as a sample of application.