DEMAND MANAGEMENT AND FORECASTING

1
DEMAND MANAGEMENT AND FORECASTING

• Demand Management and Forecasting Basics
• Short-Term Forecasting Techniques

2
CORE MRP II
MANUFACTURING RESOURCE PLANNING
3
DEMAND MANAGEMENT AND FORECASTING BASICS

• Forecasts versus orders
• Aggregate versus disaggregate forecasts
• Forecast evaluation MAD versus Bias
• Forecast techniques causal versus time series
• Forecast (time series) components

4
FORECAST VERSUS ORDERS

• Forecasts
• Are always wrong
• Become less accurate farther out
• Are necessary if manufacturing lead time is long
  relative to sales lead time
• Confirmed orders
• Are (almost) always right
• May arrive too late to meet customer's desired
  delivery date

5
AGGREGATE VERSUS DISAGGREGATE FORECASTS

• Sales forecasts can be aggregated with respect to
  time

6
AGGREGATE VERSUS DISAGGREGATE FORECASTS

• Sales forecasts can be aggregated with respect to
  time

7
AGGREGATE VERSUS DISAGGREGATE FORECASTS

• Sales forecasts can be aggregated with respect to
  product

8
AGGREGATE VERSUS DISAGGREGATE FORECASTS

• Sales forecasts can be aggregated with respect to
  product

9
AGGREGATE VERSUS DISAGGREGATE FORECASTS

• Why? Aggregate forecasts are more accurate than
  individual detailed forecasts

10
AGGREGATE VERSUS DISAGGREGATE FORECASTS

• Why? Aggregate forecasts are more accurate than
  individual detailed forecasts

11
AGGREGATE VERSUS DISAGGREGATE FORECASTS

• Use aggregate forecasts for planning long-range
  overall production levels.
• High long-range forecast accuracy
• Detail not needed for planning long-range
  resource use (labor, inventory, etc.)
• Use detailed forecasts for initial detailed
  medium-range Master Production Schedule (MPS)
• Detailed forecasts reasonably accurate for this
  time frame
• Need product-specific detail for MPS

12
AGGREGATE VERSUS DISAGGREGATE FORECASTS

• Make final short-range changes to MPS based on
  latest orders
• 100 accuracy

13
FORECAST EVALUATION

• Need to compare forecast to actual sales
• BIAS
• Too high or to low (on average)?
• Mean Absolute Deviation (MAD)
• Off by how much, per period (on average)?
• Mean Absolute Percentage Error (MAPE)
• A relative measure of MAD

14
FORECAST EVALUATION

• Need to introduce terminology

15
FORECAST EVALUATION
16
FORECAST EVALUATION
17
FORECAST EVALUATION
18
FORECAST TECHNIQUES CAUSAL VERSUS TIME SERIES

• Causal techniques
• Assume sales are determined by other variables
  which can be tracked
• Regression
• Econometric
•  Time series (autoprojection) techniques
• Make no assumption about what causes sales
• Identify patterns in past data
• Project those patterns into the future

19
TIME SERIES COMPONENTS

• Predictable components
• Base level
• Trend
• Seasonality
• Unpredictable components
• Randomness
• All time series techniques work by
• Smoothing randomness out of the past data
•  Projecting the leftover pattern implied by these
  predictable components into the future