Optimizing Bootstrap method to improve forecasting accuracy in business jet spare parts supply chains

Having products available when customer wants them keeps customers satisfied and businesses more competitive, especially in business aircrafts industry where clients are paying considerably higher for flying than in a commercial airliner. So, all parts needed for normal operation and maintenance must be readily available to ensure business flights. As there are business-aircrafts' spare parts which have lead times of up to three years, to ensure that the right parts are available at the right time in the right volume and at the right location, forecasts must be made of anticipated customer demand. The overestimated demand results in holding costs as storage, rent utilities and product obsolescence costs whereas underestimated demand leads to back order and stock out. The demand in business aircrafts industry is volatile so traditional forecasting methods are not efficient. This give rise of interest in data driven forecasting methods e.g. Bootstrap, Neural Network etc. have compared existing traditional and data driven methods and proposed an extension to Bootstrap through sliding window concept as best performing forecasting method. In this paper, optimization methodology for the sliding window bootstrap is proposed to improve forecasting accuracy. Results are validated using data from Dassault Falcon Business Jet. USA, world renowned Business Aircraft manufacturer.