Operations and Production Management MGMT 405 Answer set 3 MGMT 405 Operations and Production Management Answer set 3 (Reference chapter 3– William J. Stevenson-2007, ninth edition) Problems and Solutions 1. A commercial bakery has recorded sales in dozen for the products as shown below: Blueberry Cinnamon Muffins buns 1 30 18 45 2 34 17 26 3 32 19 27 4 34 19 23 5 35 22 22 6 30 23 48 12 31 28 26 13 35 29 27 14 34 31 24 15 33 33 22 Day Cupcakes (a) Predict orders for the following day for each of the products using an appropriate naïve method. Also Plot each data set. (b) What should the use of sales data instead of demand imply?? Ans: © 2010/11, Sami Fethi, EMU, All Right Reserved, McGraw-Hill, 2007, 9. Ed. 1 Operations and Production Management MGMT 405 Answer set 3 (a) We need to use naive methods. This means that any simple appropriate method can conducted. For example: Ploting each data set reveals that muffins orders are almost stable, varying around an average (e.g. 33). The demand for cinnamon buns has a trend. If we get the last three period, we realized that number increased two by two so 33-31=2 and last one is 33+2=35. This is for the following day for cinnamon buns. Demand for cupcakes has an apparent seasonal variation with peaks every five days. Day 1=45, Day 6=48, Day 11=47 and the next peak would be Day 16=50. A commercial bakery 60 sales 50 40 Blueberry Muffins 30 Cinnamon buns 20 Cupcakes 10 0 1 2 3 4 5 6 12 13 14 15 Days (b) The use of sales data instead of demand implies that sales adequately reflect demand. We are assuming that n stock-outs because demand equals sales if there are no shortages. 2. A company records indicates that monthly sales for a seven-month period are as follows:. Month Sales (000, unit) Feb 19 Mar 18 Apr 15 © 2010/11, Sami Fethi, EMU, All Right Reserved, McGraw-Hill, 2007, 9. Ed. 2 Operations and Production Management MGMT 405 May 20 Jun 18 July 22 Aug 20 Answer set 3 (a) Plot the monthly data as can be seen in the table above. (b) Forecast the monthly sales using linear trend equation, (c) A five-month moving average and exponential Smoothing technique? Assume that smoothing constant and march forecast value are 0.20 and 19 respectively. (d) The naïve approach (e) A weight average method conducting 0.60 for August, 0.30 for July, and 0.10 for June. (f) Which method seems least appropriate? Why? Ans: (a) Sales 25 000 20 15 Sales 10 5 0 Feb Mar Apr May Jun July Aug month © 2010/11, Sami Fethi, EMU, All Right Reserved, McGraw-Hill, 2007, 9. Ed. 3 Operations and Production Management MGMT 405 Answer set 3 (b) T (or x) Square of t Y tY 1 1 19 19 2 4 18 36 3 9 15 45 4 16 20 80 5 25 18 90 6 36 22 132 7 49 20 140 Sum= 28 Sum of square= 140 Sum= 132 Sum= 542 n= 7 x y- x xy n x -( x) 2 a= 2 b= 2 n xy- x y n x 2 -( x)2 Or If formula b is used first, it may be used formula a in the following format: n xy x y n x 2 x 2 a Y b X n b 7(542) 28(132) 0.50 7(140) 28(28) a 132 0.50(50) 16.86 7 Y = 16.86 + 0.50X After Aug (7), For sept: Y = 16.86 + 0.50(8)=20.86 (b) MA5 15 18 22 20 19 5 © 2010/11, Sami Fethi, EMU, All Right Reserved, McGraw-Hill, 2007, 9. Ed. 4 Operations and Production Management MGMT 405 Answer set 3 (c) Ft = Ft-1 + (At-1 - Ft-1) Ft = forecast for period t Ft-1 = forecast for the previous period = smoothing constant At-1 = actual data for the previous period April-F=19+0.20(18-19)=18.8 May- F=18.8+0.20(15-18.8)=18.04 June- F=18.04+0.20(20-18.04)=18.43 July- F=18.43+0.20(18-18.43)=18.34 Aug- F=18.34+0.20(22-18.34) =19.07 Sep- F=19.07+0.20(20-19.07) =19.26 (d) It may be around 20 (19.26) (e) WA=0.60 (20) + 0.30 (22) + 0.10 (18) =20.4 (f) Probably trend method because the data appear to vary around Y = 16.86 + 0.50(4) =18.86. 3. A cosmetics manufacturer’s marketing department has developed a linear trend equation that can be used to predict annual sales of its popular hand-foot cream. Ft = 80 + 15 t Where Ft =Annual sales (000 bottles), and assume that t= 0 in1990 (a) Are annual sales increasing or decreasing? By how much. © 2010/11, Sami Fethi, EMU, All Right Reserved, McGraw-Hill, 2007, 9. Ed. 5 Operations and Production Management MGMT 405 Answer set 3 (b) Predict annual sales for the year 2006 using the equation. (c ) Predict changes in annual sales the year between 1995 and 2005 using the equation. Ans: (a) Ft = 80 + 15 t 15 is the slope of the equation so this value show us whether annual sales increase or decrease. The sign of the value is positive that’s why annual sales are increasing by 15000 bottles per year. (b) Ft = 80 + 15 t= 80+ 15(16) = 320000 Where t=2006-1990=16 (c) Ft = 80 + 15 t= 80+ 15(5) = 155000 in 1995 Ft = 80 + 15 t= 80+ 15(15) = 305000 in 2005 Change in sales between the relevant years= 305000-155000=150000 4. New Car sales with monthly indexes (seasonal relatives) for a dealer for the past years are shown in the following table: Month Unit sold index Month Unit sold index Jan 640 0.8 Jul 765 0.9 Feb. 648 0.8 Aug 805 1.15 Mar 630 0.7 Sept 840 1.20 © 2010/11, Sami Fethi, EMU, All Right Reserved, McGraw-Hill, 2007, 9. Ed. 6 Operations and Production Management MGMT 405 Answer set 3 Apr 761 0.94 Oct 828 1.20 May 735 0.89 Nov 840 1.25 Jun 850 1.00 Dec 800 1.25 (a) Plot the data. Does there seem to be a trend? sales volume Unit Sales 900 800 700 600 500 400 300 200 100 0 Unit sold index Jan Feb. Mar Apr May Jun Jul Aug Sept Oct Nov Dec Month It seems that there exists a trend starting from 600 to 800. (b) Compute the Deseasonalized car sales Month Unit sold index Deseasonalize =actual sales/seasonal relative Jan 640 0.8 640/0.8=800 Feb. 648 0.8 810 Mar 630 0.7 900 Apr 761 0.94 809.6 May 735 0.89 825.8 © 2010/11, Sami Fethi, EMU, All Right Reserved, McGraw-Hill, 2007, 9. Ed. 7 Operations and Production Management MGMT 405 Jun 850 1.00 850 Jul 765 0.9 850 Aug 805 1.15 700 Sept 840 1.20 700 Oct 828 1.20 690 Nov 840 1.25 672 Dec 800 1.25 640 Answer set 3 (c ) Plot both seasonalized and Deseasonalized car sales data on the same graph. Briefly explain. sales 1000 volume 800 seasonal 600 index 400 deseasonal 200 0 Jan Feb. Mar Apr May Jun Jul Aug Sept Oct Nov Dec Month Seasonalized car sales data are upward trend in sales and Deseasonalized car sales data are downward trend in sales. 5. Two different forecast techniques (F1 and F2) were used to forecast demand for cases of bottled water. Actual demand the two sets of forecasts are as follows: Period and Demand Predicted Demand Period Unit sold F1 F2 1 68 66 66 © 2010/11, Sami Fethi, EMU, All Right Reserved, McGraw-Hill, 2007, 9. Ed. 8 Operations and Production Management MGMT 405 2 75 68 68 3 70 72 70 4 74 71 72 5 69 72 74 6 72 70 76 7 80 71 78 8 78 74 80 Answer set 3 (a) Compute MAD for each set of forecast. Which forecast is better or appears to be more accurate? Briefly explain. (b) Compute MES for each set of forecast. Which forecast is better or appears to be more accurate? Briefly explain. (c) Compute MAPE for each set of forecast. Which forecast is better or appears to be more accurate? Briefly explain. (d) Compute a tracking signal for the 8th month for each forecast using the cumulative error for 1 and 8. (e) Compute 2s control limits for each forecast. Ans: Period Demand F1 e e e2 ( e /actual)*100 1 68 66 2 2 4 3 2 75 68 7 7 49 9.3 3 70 72 -2 2 4 2.8 4 74 71 3 3 9 4 5 69 72 -3 3 9 4.3 6 72 70 2 2 4 2.7 7 80 71 9 9 81 11.2 8 78 74 4 4 16 5.1 32 176 42.4 586 © 2010/11, Sami Fethi, EMU, All Right Reserved, McGraw-Hill, 2007, 9. Ed. 9 Operations and Production Management Period Demand F2 MGMT 405 e e Answer set 3 e2 ( e /actual)*10 0 1 68 66 2 2 4 3 2 75 68 7 7 49 10.2 3 70 70 0 0 0 0 4 74 72 2 2 4 2.7 5 69 74 -5 5 25 6.7 6 72 76 -4 4 16 5.2 7 80 78 2 2 4 2.5 8 78 80 -2 2 4 2.5 24 106 32.9 (a) Compute MAD for each set of forecast. Which forecast is better or appears to be more accurate? Briefly explain. Ans: MAD= ∑│Actual-Forecast│/n MADF1= 32/8= 4 © 2010/11, Sami Fethi, EMU, All Right Reserved, McGraw-Hill, 2007, 9. Ed. 10 Operations and Production Management MGMT 405 Answer set 3 MADF2= 24/8= 3 F2 forecast results appears to be more accurate because MAD model gives less error than F1. (b) Compute MES for each set of forecast. Which forecast is better or appears to be more accurate? Briefly explain. Ans: MES= ∑│Actual-Forecast│2/ (n-1) MESF1= 176/7= 25.14 MESF2= 106/7= 15.14 F2 forecast results appears to be more accurate because MES model gives less error than F1. (c) Compute MAPE for each set of forecast. Which forecast is better or appears to be more accurate? Briefly explain. Ans: MAPE= (∑ (│Actual-Forecast│/actual)*100)/n MAPE F1=42.4/8=5.34 MAPE F2==32.9/8=4.11 F2 forecast results appears to be more accurate because MAPE model gives less error than F1. (d) Compute a tracking signal for the 8th month for each forecast using the cumulative error as well as limits of ± 4 for 1 and 8. © 2010/11, Sami Fethi, EMU, All Right Reserved, McGraw-Hill, 2007, 9. Ed. 11 Operations and Production Management MGMT 405 Answer set 3 Ans: Tracking Signalt = ∑ (Actual-Forecast)/MADt MSE Control Limits= 0 +/- 2 CUM Period Demand F1 e e 1 68 66 2 2 2 2 75 68 7 7 9 3 70 72 -2 2 7 4 74 71 3 3 10 5 69 72 -3 3 7 6 72 70 2 2 9 7 80 71 9 9 18 8 78 74 4 4 22 ERROR CUM Period Demand F2 e e 1 68 66 2 2 2 2 75 68 7 7 9 3 70 70 0 0 9 4 74 72 2 2 11 5 69 74 -5 5 6 6 72 76 -4 4 2 7 80 78 2 2 4 8 78 80 -2 2 2 ERROR MADF1= 32/8= 4 © 2010/11, Sami Fethi, EMU, All Right Reserved, McGraw-Hill, 2007, 9. Ed. 12 Operations and Production Management MGMT 405 Answer set 3 MADF2= 24/8= 3 MESF1= 176/7= 25.14 MESF2= 106/7= 15.14 MAPE F1=42.4/8=5.34 MAPE F2==32.9/8=4.11 Tracking Signalt = ∑ (Actual-Forecast)/MADt =cumulative error/ MADt Tracking Signalt for F1= 22/4=5.5 Tracking Signalt for F2= 2/3=0.66 Tracking Signal is used to monitor a forecast whether the method is biased or not. Values can be positive or negative and zero value would be ideal. Since 5.5 >4 the forecast method 1 is biased. This means that using F1, we are underestimating (positive bias)* demand. Since 0.66 < 4, the forecast is in control. *overestimating (negative bias). (e) Compute 2s control limits for each forecast. Ans: Control Limits= 0 +/- 2 MSE Control Limits for F1 = 0 +/- 2 MSE =0 +/- 2 √(25.14=+10.01 and -10.01 Control Limits for F2 = 0 +/- 2 MSE =0 +/- 2 √(15.14=+7.78 and -7.78 Since all errors for both forecasts 1 and 2 are within these limits, the forecasts are in control (see column e for both F1 and F2 in the tables above). © 2010/11, Sami Fethi, EMU, All Right Reserved, McGraw-Hill, 2007, 9. Ed. 13