IBM Maximo MRO Inventory Optimization - How is Forecast Usage Value calculated?

Answer

Forecast usage calculation is complex and influenced by multiple factors. For items with a moving code of 2 or higher, if the forecast method is set to BEST, a combination of the most suitable forecasting techniques is evaluated. These may include moving average, single exponential smoothing, double exponential smoothing, demand interval analysis, and least squares regression, each tested across various parameters.

To ensure accuracy, forecast error is calculated, and the top four methods with the lowest error margins are identified. These selected methods are then weighted and combined to generate the final forecast. Additionally, lead time is factored into the calculation, as variations over this period can significantly impact forecasting, especially when lead times are extended.

The amount of historical usage data considered depends on the item's moving frequency.

• Moving codes 0, 1, and 2: 60 months of usage data
• Moving code 3: 48 months of usage data
• Moving codes 4, 5, and 6: Up to 24 months of usage data

This tiered approach ensures that the forecast remains relevant while accommodating the item's movement patterns.