In the regeneration process of complex capital goods, the definite workload is uncertain until the goods are disassembled and inspected. Due to the uncertainty and long repair lead times, regeneration service providers have difficulties in achieving low regeneration times and meeting delivery dates. Delays in delivery are associated with contractual penalties and keeping a high stock level of spare parts coincides with a high capital tie-up. Therefore, this paper deals with the use of prognostic data mining for long-term material disposition and scheduling to accomplish a high delivery date reliability and low stock levels.