Operational Management - Delta Plastics Inc and Great Northwest Outdoor Company - Research Paper Example

Table of Contents Portfolio Exercise Delta Plastics Inc Portfolio Exercise 2: Great Northwest Outdoor Company''''''''''''..''''7 Portfolio Exercise 3: Professional Video Management'''''''''''''''.'12 Portfolio Exercise 4: Newmarket International Manufacturing ''.'''''''''...'15 References'''''''''''''''''''''''''''''''''''.... 20 Portfolio Exercise 1 Delta Plastics Inc. The defect rate is a very important facet of the manufacturing process of a product that should be taken into account. This is one of the factors that determine if the said process is still under control or not. For this purpose, there are a number of statistical methods that can be employed that provide a means of measuring how controlled a process is. A new material, "super plastic", has recently been employed as the material used in production. With the introduction of this new material, it is important to have a measure on its consistency in producing quality products. Particularly, defects in the form of uneven edges, cracks, scratches, air bubbles, and thickness variations are analyzed to reveal any problems with the new process. Tables 1 and 2 present the number of defects for the new material, "super plastic", as well as the standard material. Table 1. Defect Data for the Standard Material Standard Material Uneven Edges Cracks Scratches Air Bubbles Thickness Variation Daily Total Week 1 M 1 2 3 4 1 11 T 2 3 1 2 0 8 W 2 2 2 2 4 12 Th 3 3 3 3 0 12 F 2 0 4 4 2 12 Week 2 M 3 3 2 4 0 12 T 1 2 1 3 1 8 W 1 2 0 2 1 6 Th 2 1 2 4 2 11 F 0 3 3 3 0 9 Week 3 M 2 3 3 3 1 12 T 2 2 1 1 1 7 W 1 2 0 2 3 8 Th 3 0 1 2 0 6 F 2 2 3 4 2 13 Week 4 M 3 1 3 2 2 11 T 1 2 1 3 1 8 W 1 2 0 2 1 6 Th 2 3 1 4 2 12 F 2 3 3 1 0 9 Type Total 36 41 37 55 24 193 Table 2. Defect Data for the "Super Plastic" New Material Uneven Edges Cracks Scratches Air Bubbles Thickness Variation Daily Total Week 1 M 2 6 0 2 1 11 T 2 6 1 0 0 9 W 3 4 0 2 2 11 Th 2 3 1 1 1 8 F 0 7 0 3 2 12 Week 2 M 3 4 0 4 0 11 T 1 4 1 3 1 10 W 1 4 0 2 2 9 Th 2 3 2 4 2 13 F 3 3 2 3 0 11 Week 3 M 1 4 0 4 1 10 T 2 6 1 5 0 14 W 2 4 2 5 4 17 Th 2 4 1 5 0 12 F 3 3 0 3 1 10 Week 4 M 3 5 1 6 0 15 T 2 7 1 5 1 16 W 1 6 0 4 1 12 Th 2 3 2 6 2 15 F 0 7 3 5 0 15 Type Total 37 93 18 72 21 241 Since only raw defect data are available, the c-chart would be the best statistical tool to aid in the analysis of the new process. Since we are interested in identifying the reliability of the new manufacturing process, we take all the defects in a day as one figure. Taking the mean of this, the upper control limit (UCL) and the lower control limit (LCL) may be determined using the following formulas. Using these values, control charts may then be drawn for both the standard material and the "super plastic". These are shown in Figures 1 and 2, respectively. Figure 1. Control Chart for the Total Defects in the Standard Material Figure 2. Control Chart for the Total Defects in the New Material The two charts show that the total defects for both materials are within the computed limits. For the standard material, the number of defects fluctuates around the mean, but there is no apparent trend and there appears to be nothing suspicious with regards to the defect rate. For the new material, however, there appears to be a slightly increasing trend in the daily defect values. During the earlier days, the values were generally below the mean, and in the later days, the values have become generally greater than the mean. This suggests the possibility of the defect count exceeding the limit in the future. This may be due to the equipment used on the new material, or it could be because of the new material itself. It therefore becomes necessary to construct individual control charts for each defect type to attempt to isolate the problem to a particular process. Calculating the means of each defect type, the upper and lower control limits presented in Tables 3 and 4 may be obtained. Table 3. The Upper and Lower Control Limits for the Standard Material Standard Material Uneven Edges Cracks Scratches Air Bubbles Thickness Variation Daily Total Upper Limit 5.824922 6.345346 5.930441 7.724937 4.486335 18.96933 Lower Limit -2.22492 -2.24535 -2.23044 -2.22494 -2.08634 0.330665 Table 4. The Upper and Lower Control Limits on the New Material New Material Uneven Edges Cracks Scratches Air Bubbles Thickness Variation Daily Total Upper Limit 5.930441 11.11916 3.74605 9.2921 4.124085 22.46393 Lower Limit -2.23044 -1.81916 -1.94605 -2.0921 -2.02409 1.636067 Again, using this data, control charts can then easily be constructed. Figure 3 presents a comparative between the defect control charts of the existing standard material and the newly adopted "super plastic". Figure 3. Control Charts of the Standard and the New Material for Each Defect Type Inspecting each pair of control charts, we can compare the performance of the two processes. It can be seen that the occurrence of uneven edges follow the same trend for both processes. The same applies to the variations in the thickness of the material. However, the number of cracks in the "super plastic" has grown as compared to the standard material. It should be noted, though, that the said defect is well within the control limits and hence the occurrence of cracks is within control. For the company, it may be accepted as a difference between the two processes and is no cause for concern. Also, the number of scratches has significantly been reduced in the new material which is a rather interesting side-effect of the new process. Perhaps the most critical feature which is seen in Figure 3 is the trend of the occurrence of air bubbles. Like the daily defect trend, the occurrence of air bubbles appears to increase through time. This pattern seems to indicate that eventually, the number of defects will rise beyond the upper control limit. This is a major cause of concern for the new "super plastic" as it could hinder future production. The processes and equipment which involve the production of these defects must then be inspected to ensure that everything is in order. Having identified this problem, management can then find ways to correct the known issues. Portfolio Exercise 2 Great Northwest Outdoor Company The Great Northwest Outdoor Company is a specialty clothes store for outdoor recreational clothing. Due to the changes in seasons, the sales output of this company also varies greatly throughout the year. The data for the past five years show some interesting trends of the company's performance. This raw information is presented below. Figure 1. Historical Seasonal Demand for 2000-2004 For one thing, it is clear from this figure that the sales of the company are indeed seasonal with a peak in the fourth quarter of the year. Also, while it may be less obvious, an increasing trend may be seen through the years. For the company to effectively forecast the demand in the succeeding years, it is therefore necessary to take into consideration these two factors. The existing forecasting mechanism being employed is exponential smoothing. While this technique is quite reliable for some companies, it lacks some crucial things that are important for Great Northwest Outdoor Company. The most important of these limitations is the lack of seasonal considerations. However, it is easy to adapt the existing exponential smoothing technique to a seasonal setting. The use of individual forecasts for each season allows this technique to work effectively. This method which is currently in use can effectively model complex patterns for each season. To measure the performance of this forecasting technique, the predictions were made for 2000 to 2005. Table 1. Exponentially Smoothed Forecast for 2000-2005 Quarter 1999 2000 2001 2002 2003 2004 2005 1st 18.5 18.5 18.53 18.401 19.6007 20.6805 21.8263 2nd 23.4 23.4 23.43 23.811 25.3077 25.9954 27.4968 3rd 20.1 20.1 20.19 19.983 20.2881 21.5217 22.1752 4th 41.5 41.5 41.62 43.024 43.7668 44.7668 47.1767 Comparing this with the actual demand listed under Table 1, we obtain the absolute deviations and the mean absolute deviation for each season as shown: Figure 2. Forecasting Error of Exponential Smoothing (' = 0.3) What this error reveals is that from 2000-2001, the exponential smoothing forecast worked reliably. However, a sudden increase in error is present in the succeeding years. By cross-referencing this data with the actual demand, one can see that between 2001 and 2002, a sudden rise in sales was present leading to the failure of exponential smoothing. In fact, one of the limitations of the method itself is that it cannot effectively compensate for trends in the data. The proposed solution is to use a seasonally adjusted forecast to predict demand. The base prediction of this method is derived from using the least squares method for linear regression to obtain the best fit trend line. This is given as: y = 96.33 + 6.89x Also, seasonal indices are computed from the actual data. This is done by taking the average of each season and dividing it by the average monthly demand for the entire sample duration. This leads to an estimate of which portion of a year's demand is allocated for a given season. By obtaining an estimate of the monthly demand from the trend line and then multiplying it to the respective season, a good forecast of seasonal demand is obtained. The resulting seasonal indices along with the forecasts acquired from the 2000-2005 data are shown in the following table. Table 2. Seasonal Indices and Forecast for 2000-2005 Quarter Index 2000 2001 2002 2003 2004 2005 Total 96.33 103.22 110.11 117 123.89 130.78 1st 0.7303 17.5874 18.8454 20.1033 21.3613 22.6192 23.8772 2nd 0.9272 22.3293 23.9264 25.5235 27.1206 28.7177 30.3148 3rd 0.7453 17.9487 19.2325 20.5162 21.8 23.0838 24.3676 4th 1.5973 38.467 41.2183 43.9697 46.721 49.4724 52.2237 Again, we evaluate the performance of this technique by taking the absolute deviations and the mean absolute deviation. The resulting data is presented in Figure 3. Figure 3. Error of Seasonally Adjusted Forecast It then becomes clear that this method more closely matches the actual demand as compared to exponential smoothing. This is primarily due to the fact that seasonally adjusted forecasting takes into consideration the underlying trend. However, the method itself is not so perfect. It assumes that for each year, a constant proportion of demand between seasons is maintained. Also, the trend line itself is constrained by a linear trend which is not always the case due to changing economic conditions. These are the primary limitations of the seasonally adjusted forecast. As an alternative option, it is not always necessary to adopt a completely different system. It is possible to alter the existing exponential smoothing technique to incorporate additional degrees of freedom. One way of achieving this is to use exponential smoothing with trend. This allows the forecast to consider both the underlying trend and at the same time deal with complex fluctuations in the data. Through experimentation, a new value for ' and a second smoothing factor for the trend ' are proposed. Particularly, it was found that a value of 0.2 and 0.5 for these factors respectively minimize the resulting deviations. Shown below are the predicted demand for 2000-2005 using the alternative method followed by the absolute deviations. Table 3. Exponentially Smoothed Forecast with Trend for 2000-2005 Quarter 1999 2000 2001 2002 2003 2004 2005 1st 18.5 18.5 20.4317 21.6939 23.2042 24.4921 25.6532 2nd 23.4 23.4 25.605 27.3984 29.3834 30.673 32.0774 3rd 20.1 20.1 21.4875 22.3088 22.9849 24.0073 24.7245 4th 41.5 41.5 44.6076 47.7029 49.8789 51.5018 53.4999 Figure 4. Forecasting Error of Exponential Smoothing with Trend (' = 0.2, ' = 0.5) It is interesting to note that with this new forecasting approach, both the seasonally adjusted forecast and the seasonal exponential smoothing are outperformed. This is clearly seen in Figure 5, which presents a comparison of the mean absolute deviation for each quarter. It is therefore my recommendation that the company should adapt exponential smoothing with trend considerations. Figure 5. Comparison of MAD for Each Season Portfolio Exercise 3 Professional Video Management Steve Goodman is faced with a crucial decision with his company. His professional video system is composed of many individual modules merged into one through the use of a control box. As with all manufacturing organizations, Steve is faced with a choice of suppliers. Suppliers for a component may offer different features at different prices. Also, the availability and accessibility of the materials play their part in the choice of the preferred supplier. In the scenario, Steve already has access to the other components except the videotape system. After narrowing down a list of videotape suppliers, Steve is left to choose between two Japanese companies, Toshiki and Kony. The choice would have been simpler had the scenario been placed in an ideal setting. However, as with most suppliers, both Toshiki and Kony offer varying prices for their goods. Not only does the price vary between the companies, it also changes according to the quantity. Specific brackets are given for both organizations as to the prices. These are summarized in the table below. Table 1. Pricing Scheme of Toshiki and Kony Company' Price per Unit Toshiki 0 - 2000 2000 - 8000 8000 - 20000 20000 + 250 230 220 210 Kony 0 - 1000 1000 - 5000 5000 + ' 240 230 220 The task at hand is to determine the best of the two companies in terms of the total annual cost of ordering items. The first step to finding this optimum quantity is to apply the economic order quantity (EOQ) model on the discounted schemes. The equation for the optimum price for each quantity bracket is given as: Q* = sqrt ( 2 DS / IP ) In this equation, D is the annual demand for the product. Since the demand for the previous months have been given, the average monthly demand and the annual demand may also be computed: D = ( 7970 + 8070 + 7950 + 8010) / 4 = 8000 Knowing this, the respective optimal order quantities are shown in Table 2. Table 2. Economic Order Quantities Company' Q* Toshiki 0 - 2000 2000 - 8000 8000 - 20000 20000 + 679 2000 8000 20000 Kony 0 - 1000 1000 - 5000 5000 + ' 462 1000 5000 It is important to note that the order quantity must fall within the range of the quantity bracket. Otherwise, the lower limit is used as the best quantity for that range. Since we are dealing with quantity discounts, the data presented in Table 2 become quite incomprehensible. To obtain more relevant data, the total price is obtained from the given quantities using the following equation: TC = DS / Q + QH / 2 + PD The resulting prices for each bracket are summarized below. Table 3. Total Price for Each Quantity Bracket for Toshiki and Kony 'Company Total Cost Toshiki 0 - 2000 2000 - 8000 8000 - 20000 20000 + 48050912 44237640 42506160 40950864 Kony 0 - 1000 1000 - 5000 5000 + ' 46113255 44202180 42406536 The data presented is quite interesting. While Toshiki may appear to be disadvantageous due to its large ordering cost and its long lead time (as a result of the distance), it appears as though the lower price offering for quantities greater than 20000 units compensate for the ordering price. In fact, despite the small difference in price between Kony and Toshiki for each unit, with bulk orders, this results to a surprisingly large savings on the part of Steve. However, as we mentioned earlier, the long lead time may prove to be a problem. To fully visualize this, it is more convenient to take the reorder point of both companies as shown in the succeeding section. ROP = D x L Table 4. Reorder points for Toshiki and Kony Company ROP Toshiki 48000 Kony 8000 What these values show are quite unusual. It is impossible for the reorder point to be above the EOQ since the inventory does not even reach the needed reorder level. What this simply means is that the lead time is too long such that the demand for the products is higher than the available supply in that period. To remedy this, proper planning is necessary. The best approach Steve could use is to ensure that orders for the videotape systems are made before they are actually needed such that these goods arrive just-in-time (JIT) to meet with the demand. Also, regular and continuous reordering is necessary to ensure that the materials do indeed arrive on time. The new reorder point is then taken as the remainder of the division between the old reorder point and the EOQ. The actual reorder points for both Toshiki and Kony are shown in Table 5. Table 5. Actual Reorder Points for Toshiki and Kony Company new ROP Toshiki 8000 Kony 3000 Based on these numbers, it is safe to say that the previous recommendation of choosing Toshiki is still valid. What is important to consider is that the long lead time may lead to the use of obsolete technology and should therefore be planned and coordinated properly with the supplier. Finally, some other options Steve has on hand is to market each component of his system separately. This move will open the possibilities for several suppliers for each component. This will allow Steve to schedule the ordering process according to the individual demands and thus minimize on cost. This is possible since the assembly of the system as a whole becomes a bottleneck in the entire process. If the system is not sold, there is no need for any new parts. On the other hand, if parts of a system are sold, the demand for each is independent of the other components. Reordering in this context becomes more complicated as the individual stock levels of the components have to be monitored. However, in terms of lead time, Steve may opt to choose suppliers which offer cheap products with minimal lead time.] A final alternative for the company is to modify the control box for use with other videotape systems. Like the previous case, this will allow for greater flexibility and management of inventories. A lowered demand for the videotape system may occur which would reduce the need for large quantity orders. Depending on the actual demand, the choice of supplier may even shift from Toshiki to Kony. Likewise, the reorder point would shift depending on the demand. These alternatives offer a great deal of options for Steve in terms of cost savings and must therefore be considered carefully. Portfolio Exercise 4 Newmarket International Manufacturing (A) The recent performance review of Newmarket International Manufacturing Company (NIMCO) has revealed a few issues regarding the existing labor force. The past two quarters have revealed dissatisfactory customer service as well as production performance. It is therefore necessary to properly plan the staffing arrangement for the incoming quarter. To effectively organize this, the forecasted demands shown in Table 1 come into use. Table 1. Forecasted Demand for the Second Quarter for Each Product Line Week Demand for Product A Demand for Product B Demand for Product C 14 3600 4000 2000 15 4000 4000 2500 16 4300 4000 2800 17 4400 3800 3100 18 4500 3800 3200 19 4500 3800 3200 20 4400 3600 3200 21 4300 3600 3000 22 4000 3600 3000 23 4000 3800 2800 24 3600 3800 2800 25 3200 3800 2600 26 3000 4000 2600 As presented in the table above, the demands for each of the products appear to vary throughout the quarter. The consumer demand for Product A peaks during weeks 18-19. Product B is expected to be in demand near the start and end of the quarter however a high volume is still needed for the period in between. Product C is expected to be the least demanded of the three product lines. Based on the forecasts, the maximum demand for this product line is expected to occur around weeks 18-20. This information reveals that all product lines need to be produced more rapidly around weeks 18-19. Since the number of hours needed to complete each product is different, it is more useful to inspect the total number of work hours spent to meet with the expected demand. This is shown in the table that follows. Table 2. Forecasted Work Schedule for the Second Quarter for Each Product Line Hours for Product A Hours for Product B Hours for Product C Total Hours 864 1520 580 2964 960 1520 725 3205 1032 1520 812 3364 1056 1444 899 3399 1080 1444 928 3452 1080 1444 928 3452 1056 1368 928 3352 1032 1368 870 3270 960 1368 870 3198 960 1444 812 3216 864 1444 812 3120 768 1444 754 2966 720 1520 754 2994 This table supports our suggestion that additional labor should be placed at around weeks 18-19 while the labor force for the rest of the quarter may be relaxed. This means that the amount of labor needed for the coming quarter is not constant throughout the said period. To deal with this, three plans have been prepared and analyzed. Each of these three plans propose a method of handling the labor force such that total labor expenses should be reduced while not sacrificing customer service and the performance of the operation. The first suggestion is to use a level workforce. Since the forecasted demand is available, it is easy to determine the maximum number of employees needed based on the largest demand period. The numbers of employees needed for each week are tabulated below. Table 3. Work Force Requirement for the Second Quarter Plan A Week Employees Needed @ 87 Employees Hiring 12 More Weekly Total 14 75 48720 6000 54720 15 81 48720 0 48720 16 85 48720 0 48720 17 85 48720 0 48720 18 87 48720 0 48720 19 87 48720 0 48720 20 84 48720 0 48720 21 82 48720 0 48720 22 80 48720 0 48720 23 81 48720 0 48720 24 78 48720 0 48720 25 75 48720 0 48720 26 75 48720 0 48720 Total Labor Cost 639360 From this data, we can see that 87 employees are needed to completely deal with the peak work load. This allows us to hire additional employees and complete the production without having to deal with overtime pays. Apart from the simplicity of this plan, its appeal lies in the part where the existing workforce is maintained hence the necessary experience is also retained. This would be beneficial to maintaining the quality of the product. Furthermore, this approach can deal with fluctuations in the actual demand since excess demand can be remedied through the use of overtime while insufficient demand does not need to be dealt with. This however means that the company will incur losses during periods wherein the workforce is not maximally utilized. The total labor costs for this plan are also outlined in Table 3. An alternative to this is to simply retain the existing workforce and make use of overtime to deal with increased demand periods. This again has the advantage of using a fixed-sized workforce which is easier to deal with. Since the old workforce is maintained, the workforce maintains a level of experience with the process so it shouldn't lead to problems in terms of operations. However, the key disadvantage in pursuing this plan is that overtime pay is quite expensive amounting to 1.5 times the regular salaries. The total costs involved in this plan are summarized in the following table. Table 4. Overtime Hours for the Second Quarter Plan B Week Hours per Employee Normal Fees Excess Hours Overtime Fees Weekly Total 14 40 42000 0 0 42000 15 43 42000 3 4725 46725 16 45 42000 5 7875 49875 17 46 42000 6 9450 51450 18 47 42000 7 11025 53025 19 47 42000 7 11025 53025 20 45 42000 5 7875 49875 21 44 42000 4 6300 48300 22 43 42000 3 4725 46725 23 43 42000 3 4725 46725 24 42 42000 2 3150 45150 25 40 42000 0 0 42000 26 40 42000 0 0 42000 Total Labor Cost 616875 Comparing this approach to the first, it can be seen that it is actually more economical. It is also resilient to changes in demand. A sudden rise in the demand is simply dealt with by adding more overtime hours. While this may be expensive, it is an acceptable cost compared to losses from stockouts. On the other hand, drops in the demand would actually mean that the company saves up on overtime wages. This flexibility is the most attractive feature of this plan. The last option is to adjust the workforce for every week. This option attempts to minimize the costs by removing excess labor and adding employees when needed. In place of the overtime pay, since only regular hours are needed the only additional costs are with hiring and firing employees. As seen in the succeeding table, this amounts to a small amount which makes this plan the most appealing when it comes to financial considerations. Table 5. Workforce Adjustment Costs for the Second Quarter Plan C Week Employees Needed Hiring Firing Salary Hiring / Firing Fees Total 14 75 0 0 42000 0 42000 15 81 6 0 45360 3000 48360 16 85 4 0 47600 2000 49600 17 85 0 0 47600 0 47600 18 87 2 0 48720 1000 49720 19 87 0 0 48720 0 48720 20 84 0 3 47040 2250 49290 21 82 0 2 45920 1500 47420 22 80 0 2 44800 1500 46300 23 81 1 0 45360 500 45860 24 78 0 3 43680 2250 45930 25 75 0 3 42000 2250 44250 26 75 0 0 42000 0 42000 Total Labor Cost 607050 However, there are a few issues involved when deciding towards this plan. Since the labor force is continually resized, the amount of experience each individual has with the process is reduced. This could lead to operational problems as well as poor customer service. Also, this plan is based on the premise that the actual demand follows the forecast. Since employees cannot simply be added and removed without prior planning, adjustments in accordance to demand fluctuations are quite difficult. Weighing all these factors together, the balanced plan of choice would still be Plan B. This is since the flexibility of this plan helps in dealing with real-world circumstances. Also, the immersion of the employees in the process helps promote quality and good customer service. All these advantages are gained by trading off a relatively small difference in labor costs. It is therefore my recommendation that NIMCO should adopt this plan. References Reid, Dan and Nada Sanders. Operations Management: An Integrated Approach. 2nd. Wiley & Sons, Inc., 2005. Read More