utdallas.edu/~metin 1 Planning Demand and Supply in a Supply Chain Forecasting
utdallas.edu/~metin 2 Learning Objectives u Overview of forecasting u Forecast errors u Aggregate planning in the supply chain u Managing demand u Managing capacity
utdallas.edu/~metin 3 Phases of Supply Chain Decisions u Strategy or design:Forecast u Planning:Forecast u Operation/ExecutionActual demand u Since actual demands differ from the forecasts, … so does the execution from the plans. –E.g. Supply Chain degree plans for 40 students per year whereas the actual is ??
utdallas.edu/~metin 4 Characteristics of forecasts u Forecasts are always wrong. Include expected value and measure of error. u Long-term forecasts are less accurate than short-term forecasts. Too long term forecasts are useless: Forecast horizon –Forecasting to determine »Raw material purchases for the next week; Ericsson »Annual electricity generation capacity in TX for the next 30 years; Texas Utilities »Boat traffic intensity in the upper Mississippi until year 2100; Army Corps of Engineers u Aggregate forecasts are more accurate than disaggregate forecasts –Variance of aggregate is smaller because extremes cancel out »Two samples: {3,5} and {2,6}. Averages: 4 and 4. Totals : 8 and 8. »Variance of sample averages/totals=0 »Variance of {3,5,2,6}=5/2 –Several ways to aggregate »Products into product groups; Telecom switch boxes »Demand by location; Texas region »Demand by time; April demand
utdallas.edu/~metin 5 Forecasting Methods u Qualitative –Expert opinion »E.g. Why do you listen to Wall Street stock analysts? –What if we all listen to the same analyst? S/He becomes right! u Time Series –Static –Adaptive u Causal: Linear regression u Forecast Simulation for planning purposes
utdallas.edu/~metin 6 Components of an observation Observed demand (O) = Systematic component (S) + Random component (R) A touch of philosophy: Is the world random or everything is pre-determined? Pragmatic answer: Everything we cannot afford to study in detail is random! Level (current deseasonalized demand) Trend (growth or decline in demand) Seasonality (predictable seasonal fluctuation)
utdallas.edu/~metin 7 Time Series Forecasting Forecast demand for the next four quarters.
utdallas.edu/~metin 8 Time Series Forecasting
utdallas.edu/~metin 9 Master Production Schedule (MPS) u MPS is a schedule of future deliveries. A combination of forecasts and firm orders.
utdallas.edu/~metin 10 Summary u Aggregate forecasts are more accurate