CH-15-16 Managing Service Inventory – Service Management: Operations, Strategy, Information Technology : Book Guide

This Document Contains Chapters 15 to 16 Chapter 15 Managing Service Inventory TEACHING NOTE Inventory management is a traditional topic of the operations management course but it usually focuses exclusively on manufactured goods. However, inventory of consumable goods plays a part in services as facilitating goods (e.g., food at a restaurant). In this chapter we discuss the design of inventory management systems with applications of the latest innovations in information technology: point-of-sale scanning, electronic data interchange, bar-code tracking, and radio-frequency identification LECTURE OUTLINE 1. Inventory Theory Role of Inventory in Services (Figure 15.2) Characteristics of Inventory Systems Relevant Costs of an Inventory System (Table 15.1) 2. Order Quantity Models (Figure 15.3) Economic Order Quantity (Figures 15.4, 15.5, and Table 15.2) Inventory Model with Quantity Discounts (Figure 15.6) Inventory Model with Planned Shortages (Figure 15.7 and Table 15.3) 3. Inventory Management Under Uncertainty (Figure 15.8) 4. Inventory Control Systems Continuous Review System (Figure 15.9) Periodic Review System (Figue 15.10) The ABCs of Inventory Control (Figure 15.11 and Table 15.4) Radio Frequency Identification 5. Single-Period Model for Perishable Goods Expected Value Analysis (Table 15.5) Marginal Analysis (Figure 15.12) 6. Retail Discounting Model TOPICS FOR DISCUSSION 1. Discuss the functions of inventory for different organizations in the supply chain (i.e., manufacturing, suppliers, distributors, and retailers). Answer: Manufacturing Factory inventory can be divided into three categories: raw materials, work-in-progress (WIP), and finished goods. Raw materials usually are purchased in bulk quantities that often take advantage of price discounts or forward buying in anticipation of future price increases. A raw material stockout can be an economic disaster because the entire plant might close for lack of material to process (e.g., sand in a glass factory). Work-in-progress is material found on the factory floor in various stages of completion. WIP creates buffers between different manufacturing stages and thus allows some independence in operations. However, excessive WIP can hide quality problems and is analogous to waiting lines in services and missed delivery dates. Finished goods, if not representing a completed customer order ready for shipment, often are products produced in anticipation of future sales that are thus available for immediate delivery. Of course, this practice of stocking finished goods is the essence of the fast food concept pioneered by McDonald’s. Suppliers In addition to the role of inventory shared with other manufacturers, suppliers are expected to make deliveries to their manufacturing customers on a just in time (JIT) basis. Using electronic data interchange (EDI), suppliers are in direct contact with their manufacturing customers and thus they can reduce ordering delays significantly. Using EDI and JIT delivery, inventory levels can be reduced. Distributors Distributors act as consolidators of retail demand and provide more efficient distribution of goods, because demand can be aggregated at regional levels. Distributors allow small manufacturers the opportunity to reach markets efficiently by sharing the distribution costs with other firms. Retailers At the retail level, inventory is on display for ready access by customers. Lack of inventory for immediate sale (i.e., stockout) represents a lost sale and possible loss of future sales if the customer never returns. 2. How would one find values for inventory management costs? Answer: Ordering costs The cost of placing orders depends upon the level of information technology in use. For firms using electronic data interchange (EDI), the variable cost of placing an order is approximately zero but the investment amortization and communications costs are significant. The cost of placing an order for firms with a purchasing department should include staff salaries and overhead. However, the cost of placing an order will vary depending upon the level of purchasing activity (i.e., negotiating prices, locating vendors). Receiving and inspection costs For our inventory models the receiving and inspection costs are included in the cost of placing an order (S in the model), because these costs occur each time an order is initiated (i.e., D/Q times per year). These costs arise from activities in the warehouse and thus include labor costs for handling. Inspection costs will vary depending upon the level of trust with particular vendors. Holding or carrying costs The major component of holding costs is the opportunity cost of capital or interest charges on money borrowed (e.g., floor plan costs for automobile dealers). The investment in warehouse facilities and related variable costs of utilities and labor for stocking and retrieving inventory can also be expensive. Hidden costs of inventory include damage, theft, and depreciation (e.g., in retailing write-downs of non-moving stock). Shortage costs Shortage costs are perhaps the most difficult to estimate because lost sales or, worse, lost customers, seldom are recorded. Tangible costs such as expediting an order by overnight mail thus become minor compared to the real costs of lost customers. 3. Compare and contrast a continuous review inventory system with a periodic review inventory system. Answer: Characteristic Periodic Review Continuous Review Order Quantity Varies (Target–Inv. Level) Fixed (EOQ) Order Interval Fixed (Q*/D) Varies (Inv. Level < ROP) Inventory Monitoring Once Every Period Continuous (Using POS) Management Allows Consolidation Automatic if Computerized Safety Stock Applicability C items A items 4. Discuss how information technology can help to create a competitive advantage through inventory management? Answer: Holding inventory is a hedge against uncertainty. With perfect information on customer demand, inventory could be reduced drastically. Using point-of-sale (POS) technology, a database of customer demand by stock keeping unit (SKU) and buying frequency can be assembled to anticipate future demands and thus plan inventory-stocking decisions accordingly. For example, McDonald’s is able to estimate for a particular location over a period of time the quantity of items to prepare ahead of the lunch hour in anticipation of demand adjusted for the day and week of the year. 5. How valid are the assumptions for the simple EOQ model? Answer: The classic EOQ model, which assumes a constant rate of demand and no stockouts, is valid for inexpensive staples such as milk, bread, and gasoline that are purchased by a large number of customers in small amounts on a periodic basis. The aggregate demand of these many purchases appears as a constant rate and customers will not tolerate stockouts of these necessities as many of us saw during the gasoline crisis of the early 1970s. 6. How is a service level determined for most inventory items? Answer: Many factors are considered in determining a service level for an inventory item. The most obvious include the cost of holding extra inventory as a safety stock and the cost associated with a stockout. Using an information database and express delivery, these costs can be reduced significantly. For example, some retailers will have a stockout item shipped directly to the customer from another store. The competitive circumstances or willingness of the customer to place a backorder also play a part. Rural stores with distant competitors seldom will have the variety found in urban areas. Newsvendor 7. Service capacity (i.e., seats on an aircraft) has characteristics similar to inventories. What inventory model would apply? Answer: A distinctive characteristic of service firms is time-perishable capacity. For example, for an airline, an empty seat on a flight represents capacity lost forever. The single-period model for perishable inventory or “newsvendor problem” applies to the management of airline seat inventory. The critical fractile P(D<Q), determined using Equation 15, could be used in capacity planning. For example, the lot size Q could represent the seat capacity of the airplane. The cost of underestimating demand would represent the opportunity cost of a lost sale owing to a stockout of seats. The cost of overestimating demand represents an empty seat not sold with zero salvage value. We made use of the critical fractile model in Chapter 11 when addressing the overbooking problem and yield management. 8. Identify dependent and independent demand for an airline and a hospital. Answer: For an airline: Independent demand: passengers flown. Dependent demand: • Food and drink inventory. • Pillows and blankets. • In-flight magazines. • Size of crew (e.g., flight attendants). For a hospital: Independent demand: patients admitted. Dependent demand: • Nurse staffing. • Hospital beds for emergency cases. • Waiting-area seating. • Essential medications, such as anesthetics and “clot-busters.” • Patient gowns and staff uniforms. INTERACTIVE CLASS EXERCISE The class engages in an estimation of the cost of a 12-ounce serving of Coke in various situations (e.g., supermarket, convenience store, fast-food restaurant, sit-down restaurant, and ballpark). What explains the differences in price? This exercise illustrates the place utility of an inventory item and any associated service component. Begin with the supermarket as the lowest cost serving of Coke based on buying in quantity (6 pack) and a cost saving passed on to customers because the store achieves a price discount from the supplier by ordering in large quantities. The convenience store is able to charge a premium because the item is an impulse buy and available at many convenient locations. The fast-food restaurant charges an even higher premium because the drink is iced and is packaged with a meal. In fact, the margin on drinks subsidizes the meager profits on the hamburger. A sit-down restaurant is adding the cost of service and atmosphere to the price of drinks. Finally, an outrageous premium is charged at the ballpark because the audience is captive. EXERCISE SOLUTIONS 15.1 (a) (b) Orders per year (c) Working days between reorders 15.2 (a) Note: Avg. daily demand = 3 from lead time distribution (b) For: Q = 120 and C = 10 For: Q = 360 and C = 9.50 Therefore, take the $ .50 discount (c) Service level = 95 percent 15.3 (a) (b) (c) At 10 cubic feet per case, the refrigerated warehouse could accommodate an order quantity of only 50 cases assuming no safety stock. Dutch Farms could save $6 per year if it could rent 100 more cubic feet of warehouse space. 15.4 (a) (b) Annual Value of Discount = ($0.25)(600) = $150 Net Cost = $312 – $150 = $162 Therefore, do not take the $0.25 discount. (c) R (Reorder Point) = 4 for 85 percent service level. SS (Safety Stock) = 15.5 (a) (b) Note that in the formula above, the number of hardbacks in stock (dogs) cancels out. 15.6 15.7 (a) (b) (c) 15.8 (a) (b) 95 percent service level 15.9 (a) (b) (c) 15.10 (a) (b) Probability of stockout * 5 percent Lead time demand * uniform (30, 70) Expected lead time demand = 50 (c) (d) 15.11 (a) (b) 15.12 (a) (b) (c) Max Inventory 1149 * 104 = 1045 (d) (e) 15.13 (a) (b) (c) (d) Recommend backorder inventory policy because the backorder policy saves: $416 * 369 = $47 A maximum wait of 5 days is not unreasonable. (e) No backorder case: Backorder case: K = 8 units will be backordered Therefore, reorder whenever on-hand inventory reaches: 10 * 8 = 2 units 15.14 (a) (b) where: 15.15 15.16 (a) (b) (c) (d) Cost of coal stockout equals the cost of generating electricity with alternative fuel (natural gas) or purchase of electricity from another utility on the grid. 15.17 (a) (b) (c) (d) 15.18 (a) Using Equation 7 for calculating K*, we can solve for B because, from the problem statement, the ratio K*/Q* should equal 1/10 and thus: Solving yields B = 18 (b) CASE: A.D. SMALL CONSULTING Use inventory models to address Lou Carlson's questions. Support your recommendations with cost justification. The training program at A.D. Small is analogous to an inventory system in which the new recruits are a batch size that is processed through a training session and then placed into inventory until replacements are needed. The corresponding values for the EOQ formula are: S = 850,000, D = 180, and H = 90,000. This suggests that 3 training sessions per year should be scheduled instead of just one. The savings of $3,700,00 can be estimated using the total inventory cost function: TCp = S(D/Q) + H(Q/2). Sessions/Year Q S(D/Q) H(Q/2) TCp 1 180 850,000 8,100,000 8,950,000 3 60 2,550,000 2,700,000 5,250,000 CASE: LAST RESORT RESTATURANT 1. Assuming that the cost of stockout is the lost contribution of one dessert, how many portions of Sweet Revenge should the chef prepare each weekday? Answer: This question can be answered with the single-period model for perishable goods Equation (15). where: This assumes that the loss incurred in the stockout of one dessert is the contribution. The cost of overage is assumed to be the out-of-pocket cost to produce one additional dessert wasted if not sold. Ordering the 20 observed weeknight demands in ascending order we develop the demand distribution below with cumulative probability P(D) shown in the second row (note that two values of 250 and 275 were observed). Thus, stocking 300 deserts satisfies our critical fractile of 0.75. 2. Based on Martin Quinn’s estimate of other stockout costs, how many servings should the chef prepare? Answer: Using Equation (15) with the revised stockout value we find the need to stock 330 desserts to satisfy our critical fractile. The value 330 is an interpolation between 315 and 340. 3. If, historically, desserts were prepared to cover 95 percent of demand, what was the implied stockout cost? Answer: Using Equation (15) set equal to 0.95, we solve for the value of Cu = 19. CASE: ELYSIAN CYCLES Develop an inventory control plan for EC to use as the basis for its upcoming year. Justify your reasons for choosing a particular type (or combination of types) of inventory system(s). On the basis of your particular plan, specify the safety stock requirements if EC institutes a policy of maintaining a 95-percent service level. Answer: Because sales data is not available at the level of bicycle size (e.g. 18, 21, 23), the stock-keeping unit (SKU) will be identified by bicycle frame style (i.e., A, B, C, D, or E). It is assumed that orders for a particular frame style will be broken down into appropriate bicycle sizes based on current sales experience. Several possible inventory systems might be appropriate for Elysian Cycles. First we will consider two versions of the Continuous Review System. In the first, or disaggregate case, an inventory system complete with EOQ and Reorder point will be designed for each bicycle frame style. In the second, or aggregate case, all of the bicycles will be considered together with only one order placed for a predetermined mix of bicycle frame styles. Finally, a Periodic Review System is proposed in which the periodic order will consist of a mix of bicycle styles based on current sales. Continuous Review System (Disaggregate Version) The economic order quantity for each bicycle frame style will be calculated using equation (3) with H = IC. where: D = Annual demand for last year increased by a 15 percent growth factor S = Ordering cost of $65 I = Holding cost of 0.75 percent per month or 9 percent per year as a percentage of bicycle value C = Wholesale cost of bicycle at 60 percent of suggested list price The reorder point for each bicycle style is calculated based on the desire for a 95 percent service level. The lead-time to receive orders from the overseas manufacturer is stated as four weeks or approximately one month. It is assumed that the monthly demand data is normally distributed. From Appendix A we find that a standard normal deviate of 1.65 ensures 5 percent in one tail. Using equations (8, 9,and 11), the reorder point is calculated for each bicycle style. However, because the lead-time (LT) is approximately one month the reorder point formula is simplified as below with  and  representing the monthly demand parameters. The average monthly demand will be inflated by 15 percent to reflect anticipated growth. The standard deviation of the monthly demands will be calculated assuming dispersion from one year to the next will remain unchanged. Furthermore, a sample standard deviation is calculated using (n-1) because only 12 months of data is available. For comparison purposes the annual total cost of the inventory system is calculated using equation (2) with H = IC. Continuous Review System (Aggregate Version) Because one manufacturer furnishes all bicycles, consolidating all bicycle styles into one order saves ordering costs. Whenever an order is placed, it will contain a predetermined mix of bicycle styles. The order size of each style will be in proportion to its sales demand. The economic order quantity and reorder point is now based on the total demand for all bicycles irrespective of style. The retail cost per bicycle of $170 used above is a weighted average based on annual sales. The sample standard deviation for monthly demand used above is calculated from the last column of total monthly demand data. The number of units in the safety stock is taken directly from the previous calculations of safety stock by bicycle style. The total annual system cost is calculated as: Thus, in comparison with the disaggregate version, considerable savings in inventory system cost is achieved with some loss in controlling inventory levels by bicycle frame style. Because a fixed mix of styles is always ordered, the system will not respond to changes in demand for individual bicycle styles (possibly running out of stock of some styles and overstocking others). Periodic Review System The cost savings of aggregate ordering are retained and the ability to control inventory levels by bicycle style is achieved using the periodic review system. For the periodic review system, the stock level of each bicycle style will be reviewed and the needs for each will be aggregated into one order that can vary in quantity. The review period (RP) is based on the EOQ of 108 calculated above for the aggregate version of the continuous review system. An EOQ of 108 represents placing an order approximately every 1½ months because average aggregate monthly demand is 69. The target inventory level is determined using equation (13): For this system, an inventory of all bicycle styles is undertaken every 6 weeks. An order is placed to include the number of bicycle styles needed in each category to bring the existing inventory up to the target level for each style. The aggregate order feature is preserved, but each order mix reflects the current sales of bicycles by style. The annual inventory cost for this system is equivalent to the aggregate continuous system, because the same EOQ is used. However, because the buffer stock is now 107 units compared to 68, some additional inventory holding cost is incurred. On an annual basis this extra cost is calculated to be (.09)(107-68)(170)(.60) = $358.02. This results in a total annual system cost of $1352.10. Chapter 16 Managing Service Projects TEACHING NOTE With the emphasis on teamwork in organizations, project management skills will be expected of future leaders. Once a project manager has been selected and a team formed, management activities involve planning, scheduling, and controlling. Computer software such as Microsoft Project make these activities easier. Using this software, visual displays such as Gantt charts and PERT charts can be produced and revised with ease. Performing critical path analysis no longer is a tedious process. Project managers are freed to focus on completing the project on time and within budget. SUPPLEMENTARY MATERIALS Case: Paymor Shopping Center (Stanford University case) The owner of a shopping center is preparing for the construction of a tire store on his property. He is interested in learning how long the project will take and what activities could be expedited to complete the construction in 58 days at minimum additional cost. Software: Microsoft Project This commercial software program automatically constructs Gantt charts and project network diagrams (both using color to highlight the critical path) based on inputs of project activity times and precedence requirements. Other features include managing resources, workloads, and calendars. Controlling the project cost, scheduling task constraints, and tracking project progress are made easy. LECTURE OUTLINE 1. The Nature of Project Management Characteristics of Projects Project Management Process (Figure 16.1) Selecting the Project Manager Building the Project Team Principles of Effective Project Management 2. Techniques for Project Management Gantt Project Charts (Figure 16.2) A Critique of Gantt Charts Constructing a Project Network (Table 16.1) Critical Path Method (Figures 16.3, 16.4, 16.5, 16.6, 16.7, and Table 16.2) Microsoft Project Analysis (Figures 16.8 and 16.9) 3. Resource Constraints (Figures 16.10 and 16.11) 4. Activity Crashing (Figures 16.12, 16.13 and Tables 16.3, 16.4, 16.5) 5. Incorporating Uncertainty in Activity Times Estimating Activity Duration Distribution (Figure 16.14) Project Completion Time Distribution (Table 16.6 and Figures 16.15) A Critique of the Project Completion Time Analysis (Figure 16.16) 6. Problems With Implementing Critical Path Analysis 7. Monitoring Projects (Table 16.7) Earned Value Chart (Figure 16.17) Project Termination Project History Report TOPICS FOR DISCUSSION 1. Give an example that demonstrates the trade-off inherent in projects among cost, time, and performance. Answer: The development of a military fighter best illustrates the trade-off found in projects. The government is interested in a fighter that will not exceed budget allocations but the military desires a plane that will push the performance envelope. Achieving outstanding performance requires the use of exotic materials and results in additional expenses. Both the military and the government are interested in having the final product yesterday but accelerated development requires additional manpower and associated cost. Engineers know that with additional time the design can be improved and meet higher levels of performance. 2. Illustrate the four stages of team building from your own experience. Answer: As a young boy I played shortstop on a school baseball team. The forming stage began the first day we all showed up for tryouts with the expected excitement and anticipation, and some anxiety. After the team members were selected, positions were assigned and we took to the field for the first time; this storming stage began with the reality that we did not work well together. Fly balls were dropped, throws to first base were off the bag, and the shortstop did not cover second base when needed or flubbed an opportunity for a double play. After more practice and a few competitive games we began to relate to one another. As we entered the norming stage of our activity, we learned to anticipate each other’s moves and trust each other to field the ball. Finally, the team began to perform and play well enough that we took pride in our performance and actually won some games with appropriate celebration. 3. Are Gantt charts still viable project management tools? Explain. Answer: A Gantt chart is a useful visual representation of project activities against a time schedule with progress easily displayed. For small-scale and repetitive projects (e.g., airplane turn-around between flights), a posted Gantt chart describes for everyone what needs to be accomplished and the expected time for each activity. Gantt charts also can be used for training new personnel. 4. Explain why the PERT estimate of expected project duration always is optimistic. Can we get any feel for the magnitude of this bias? Answer: Analysis of projects with uncertain activity times proceeds with the assumption that the critical path activities based on expected activity durations actually determine the expected project completion time. However, the completion times of every path in the network are random variables. Paths whose expected durations are close to that of the critical path can, in fact, determine the project completion time owing to excessive variance in some activities. By focusing on only the critical path and ignoring other paths, the calculation of project completion time becomes an optimistic estimate. A better estimate of the project completion time distribution can be obtained using Monte Carlo simulation of the network by taking a sample of each activity time from its Beta distribution and determining the critical path. Repeated sampling will result in realizations of project durations that can be displayed as a completion time distribution. The mean of this empirical distribution then can be compared to the expected project duration (using only the critical path activities) to arrive at a measure of the bias. 5. Discuss the difference between time variance, cost variance, and schedule variance. Answer: Variance represents a measure of deviation of actual performance from planned performance. Time variance looks at work performed and compares actual with scheduled times to identify whether the project is behind or ahead of schedule. Cost variance is concerned with a comparison of the accumulation of expenses to date with budgeted cost to signal the potential of a project cost overrun in time to make adjustments, if necessary. Schedule variance is another measure of cost performance but relates budgeted cost (baseline) to budgeted cost of work performed to date. Cost variance and schedule variance will be identical only when the actual cost equals the baseline cost. However, in this case the project could still be in trouble because the value completed could be less than budgeted cost. 6. Go to http://www.people.hbs.edu/besty/projfinportal/index.htm and find employment opportunities in project finance. What role does finance play in projects? Answer: Finance plays at lease three roles in projects: (1) arranges for project financing using debt, partnerships, and/or government financing, (2) challenges the project plan for possible financial risks, and (3) challenges host government, joint venture, and supplier contracts for possible schedule and performance risks during project execution and keeps track of the costs against budget (cost variance) to avoid cost overruns. INTERACTIVE EXERCISE Prepare a work breakdown structure (WBS) for a homecoming dance. The following is a possible WBS for the homecoming dance project plan. Using WORD, the outline is numbered automatically. 1.0 Homecoming Dance 1.1 Ballroom 1.1.1 Find an available ballroom 1.1.2 Sign a contract for a date 1.2 Band 1.2.1 Decide on the type of band 1.2.2 Find an available band 1.2.3 Sign a contract for a date 1.3 Food service 1.3.1 Find a caterer 1.3.2 Sign a contract for a date 1.4 Decorations 1.4.1 Purchase balloons and streamers 1.5 Announcement 1.5.1 Print flyer 1.5.2 Post announcement in school paper 1.6 Dance day 1.6.1 Have committee show up early 1.6.2 Identify cleanup crew EXERCISE SOLUTIONS 16.1 (a) (b) Activity ES LS EF LF TS A 0 1 4 5 1 B 0 6 3 9 6 C 0 0 4 4 0 D 4 5 10 11 1 E 4 12 7 15 8 F 4 4 9 9 0 G 10 11 14 15 1 H 9 9 15 15 0 I 15 15 17 17 0 (c) Critical path activities: C, F, H, I Expected Project Completion Time: 17 days (d) A maximum of three workers is required in the resource-leveled schedule below: Critical Path Activities  Activities with Slack  (c) Critical Path Activities: B, D, G, H Expected Project Completion Time: 16 days Critical Path Activities  Activities with Slack  Critical Path Activities: C, E, F, J Expected Project Completion Time: 28 days (c) Begin team selection 28 days  3 days for tournament = 25 days before December 27. Thus, begin team selection on the morning of December 2, but no later than December 4 (2 days TS). Arranging accommodations must begin no later than December 2. 16.4 (a) (b) Using the same methodology as in part (a), we can crash the completion time of the project to 33 weeks. The actual activity time and corresponding cost is as follows: Bid price should be (1.2) (231,000) = $277,200 16.5 Maximum crash limit achieved because critical path A - E cannot be reduced further (i.e. all activities crashed to their limit) * Activity crashed to limit 16.6 (a) (d) 1. NO, activity B not on critical path 2. Crash activity G by one day yields savings of $100,000 - $50,000 = $50,000 3. Crash activity O by 2 days yields savings of $100,000(2) - $150,000 = $50,000 making N and P critical 4. Crash activity O by 2 days by drawing resources from activity N extending it by 2 days yields savings of $100,000 or one day resulting in N and P becoming critical and O non critical 16.7 (a) Critical Path Activities: B, E, H Expected Project Completion Time: 16 months (c) P(Completion Time  24) = 100% P(Completion Time  41) = .5 + .2881 = .7881 P(Completion Time  24) = .5  .1736 = .3264 16.10 (a) (c) Critical Path Activities: A, D, G, I Expected Project Completion Time: 22 weeks P(Completion Time  25) = .5  .4803 = .0197 CASE: INFO-SYSTEMS INC. [This analysis provided by Jennifer White, Ben Harrington, Hugh Ramsey, and Monica Prihoda] 1. Using Microsoft Projects, prepare a network and identify the critical path activities, the expected project duration, and scheduling times for all activities. Answer: As shown below, the critical path consists of the following activities: 2-5-6-8-9-14-20-24-28-29-30-31 with duration of 293 days. To prepare a network using Microsoft Projects, input all project activities and their dependencies to generate a visual network diagram. Identify the critical path by analyzing which sequence of activities determines the longest duration, thereby establishing the project's overall timeline and scheduling times for each task. 2. The elapsed time for delivery of the hardware is estimated at 90 days. Would the project completion time be affected if delivery of the hardware were delayed by 30 days? Would the critical path change? Answer: If the delivery of hardware were delayed by an estimated 30 days, the project completion would be postponed by 25 days. This delay can be attributed to a change in the revised critical path, as shown below, with activity 4 now taking 130 days. The critical path changes to 2-4-13-20-24-28-29-30-31, which moves the finish date forward to October 22, 1997. 3. Using the original network and critical path, what strategies could management consider to complete the project on time if activity B were delayed by several weeks? Answer: Two major types of strategies are available to complete the project on time if activity 2, “Develop batch processing system requirements,” were delayed. The first is to work weekends on the critical path activities. Returning to Microsoft Project and changing the specifications to allow work on weekends results in finishing the project in 177 days. This allows activity 2 to be delayed (293-177) or 116 days and still finish the project on time. The second option is to crash certain activities on the critical path. We can look at the end of the critical path (20-24-28-29-30-31) for opportunities. An example is to crash activity 29, “Operate system in parallel and train,” by using longer and more intense training sessions. Crashing early activities on the critical path might result in changes in the critical path. As an example of a two-part crashing strategy, we first crash activities 8 and 9 by 10 and 15 days, respectively. Crashing could be accomplished by having employees work overtime, outsourcing, or diverting employees from non-critical activities. Once these are crashed, a new critical path that includes activity 4 develops. We then crash activity 4. This can be done by paying more money to complete software earlier and by using overnight shipping. Other possibilities could be considered, but working weekends and crashing selected critical path activities should be the focus for reducing project duration. CASE: WHITTIER COUNTY HOSPITAL [This analysis provided by Jason Cain, Lydia Radnik, and Suzanne Schmidtz] 1. Assume that you are part of the management staff whose task is to develop this sketch plan. Using Microsoft Project for Windows, develop the PERT network as outlined above, identify the critical path, and determine the expected time to reach basic operational status at the new facility. Answer: Expected time to reach basic operational status at the new facility is 62.67 days. Using Microsoft Project, create a PERT network by inputting all project activities, their durations, and dependencies. Identify the critical path by analyzing the longest sequence of dependent tasks, and determine the expected time to achieve basic operational status at the new facility based on the critical path duration. 2. The board of directors has said that it would like to try to move on a Sunday to minimize interference with weekday traffic. If there are Sundays that fall 46, 53, 60, 67, and 74 days from now, determine the probability (using a normal distribution) of reaching basic operational status at the new location on the two Sundays that are closest to the expected time you calculated previously. Answer: Assuming that Whittier County Hospital starts its move planning at 8:00 a.m. on Monday, June 16, 1997, and work continues during the weekends, then the project will take an expected duration of 60.55 days finishing at approximately noon on Friday, August 15, 1997. The non-critical equipment move has been excluded from the critical path because we are interested only in the time to reach basic operational status at the new facility. If there are Sundays falling on the 60th and 67th days from now, then the probabilities of reaching basic operational status at the new location on one of these two Sundays is the following: Scenario A: Probability of finishing before Sunday, the 60th day from now P(Project Duration  60) = 0.5000 - 0.2852 = 0.2148 or approximately 21 percent. Scenario B: Probability of finishing before Sunday, the 67th day from now P(Project Duration  67) = 0.5000 + 0.4713 = 0.9713 or approximately 97 percent. 3. Briefly assess the potential problems you see in applying critical path analysis to the sketch plan for moving Whittier County Hospital. Answer: Activity 2, Completion of Construction, accounts for 64 percent (50/78) of the expected project duration. This activity has a range of 40 to 60 days that accounts for 78 percent (11.11/14.29) of the variance in project duration. Thus, the project is dominated by one very time-consuming activity that requires close scrutiny. Activity 19, Move the Patients, could have unforeseen problems, because this final activity is very delicate and could involve critically ill patients admitted to the hospital during the approximately 2 ½ month project. Applying critical path analysis to the sketch plan may face problems such as inaccurate activity duration estimates and unforeseen dependencies, which can lead to delays and misalignment with the actual project timeline. Additionally, changes in project scope or resources may impact the critical path and overall scheduling. Solution Manual for Service Management: Operations, Strategy, Information Technology James A. Fitzsimmons, Mona J. Fitzsimmons 9789339204471