Control Spare Parts Inventory Obsolescence by Predictive Modelling

The Factory supply chain management spends billions of dollars every year to procure and manage equipment spare parts. Due to technology and processes changes, some of these spares become obsolete or dead inventory. Factories have huge dead inventory worth millions of dollars accumulating over time. This is due to lack of a scientific methodology to identify them and send the inventory back to the suppliers on a timely basis. The standard approach followed across industries to deal with this is to declare a part as dead if a part is not used for a set pre-defined period of time. This leads to accumulation of the dead parts over time and these parts cannot be sold back to the suppliers as it is too late as per contract agreement. Our main idea is that the time period for identifying a part as dead cannot be a fixed pre-defined duration across all parts. This time period should depend on various properties of the part like the historical consumption pattern of the part, type of part, how many machines this part is associated with, if this part is a preventive maintenance part etc. We have designed a predictive algorithm which predicts part obsolescence well in advance with reasonable accuracy thus helping save millions of dollars.