Cost estimation for remanufacture with limited and uncertain information using case based reasoning

Abstract

Assessing products prior to remanufacture is an important part of the remanufacturing process, ensuring that unsuitable cores are removed at an early stage to avoid unnecessary processing. In particular, understanding the economic cost of remanufacturing a product can be an important aspect of the assessment, especially for businesses operating in low volumes and on high value products, where the risk associated with unexpected costs or failure to complete remanufacture are much greater. Estimating these costs can however be difficult, as important information required to make a prediction is often uncertain, such as the product design, its condition and also the understanding of the resource requirements for remanufacture. Within this research a method has been developed to estimate the economic cost and risks of conducting a remanufacturing activity to a product when information is uncertain. Summation of the individual activities can then be conducted to determine the economic cost and risks of the entire remanufacturing process. The method utilises a combination of case based reasoning and probability theory to identify similarities between historical data records and the product under assessment, to predict the cost and risks of remanufacture. In particular this method enables cost estimation when important product information is missing including the manufacturer, model or condition. Additionally estimates can be made when exact historical information is not present, which can be useful to business remanufacturing bespoke or rare products. The method is then implemented within a service oriented architecture and functionally demonstrated using an example of an independent wind turbine gearbox remanufacturer.