Simulation Project: Improving the X-Ray Process at County Hospital County Hospital wishes to improve the service level of its regular Xray operation, which runs from 8 A.M. to 5 P.M. Patients have identified the total required time as their main concern with this process. Management, on the other hand, is concerned with utilization of available resources. Management has created a process-improvement team to study this problem. The process might be redesigned as a result of the team’s recommendations. The team has defined the entry point to the X-ray process to be the instant a patient leaves the physician’s office en route to the X-ray lab. The exit point has been defined as the instant at which the patient exits the X-ray facility and the completed X-ray is delivered to the physician’s office. Patients are registered as they enter by a desk assistant, and a sample of the arrival time data is provided in Table-1. Until now, no attempt has been made to further analyze this data, so there is no insight into what the arrival process looks like. The team has identified 9 activities in the current X-ray process (see Table-2), which is the same irrespective of the patient. The activity times and their distributions are specified in Table-3. The resource data for the X-ray process is specified in Table-3. There are one desk clerk, three X-ray technicians, one dark room technician, two X-ray labs, one dark room and one X-ray inspection room. The two X-ray labs and the inspection room are staffed by Xray technicians while the dark room is staffed by a dark room technician. PART I: ANALYZING THE CURRENT PROCESS DESIGN 1. Step #1--Draw a flowchart of the current X-ray process. 2. Step #2--Develop a simulation model of this process.  The model requires analysis of input data regarding the arrival process of patients. Look at the lecturePowerPoint materials on “analyzing input data” as a guide for your modeling. A combination of descriptive statistics and a histogram can help you “eyeball” the data to help determine the most appropriate probability distribution function to help model interarrival times.  Modeling Hint: Build the model incrementally based on your flowchart. Do not try to put everything together at once and then test whether it works. The key is to use individual, small steps.  Be sure to specify your priorities and resources correctly with SimQuick. Remember, a higher priority has a lower number workstation. For instance, if you have two workstations that compete for a resource, the workstation with the higher priority is always entered into the model first.  Table-1 presents patient arrival times. You need to come up with some estimate of interarrival times, meaning the time between arrivals, to come up with a pattern for how patients enter into the X-ray facility. I highly encourage you to create a histogram, run the descriptive statistics from the analysis toolpack in Excel and estimate an appropriate, average interarrival time.  You may need to include more buffers than noted in Table-2. Assume that there’s plenty of space available for patients to wait while they’re in the service system. 3. Step #3—First cut analysis based on one simulation run: As a first-cut analysis, run one simulation, using the correct activity time distributions. Look at the average cycle time, the throughput rate, the queue lengths, and the descriptive statistics, such as the mean and standard deviations. What are the problems in this process? 4. Step #4—Multiple simulation runs: Run 100 simulations and compute the cycle time and daily throughput (average and standard deviation). Also compute the queue and resource utilization statistics with 95 percent confidence intervals. Assume that any patients remaining in the system at the end of the day will be taken care of by the night shift. Every morning, the system is assumed to be empty. Are there any surprises when you compare these results with the ones inquestion 3? 5. Step #5—Analysis based on multiple simulation runs: Assess the performance of the process using the values calculated in question 4. Where is the bottleneck? What is the problem area in terms of reducing the cycle time and increasing the throughput rate? PART II: SUGGEST AND EVALUATE NEW PROCESS DESIGNS 6. Step #6—Creative process design based on acquired process understanding: Based on your insight about the current operations, identify and model at least two plausible ways of reducing the average cycle time by redesigning the process. For example, what if more personnel are hired? What type of personnel would be most useful? What, if any, improvement would adding a dark room and/or an X-ray lab have on cycle time and throughput? What if the X-ray technicians receive training designed to reduce the probability of rejecting X-rays from 25% to 10%? Test at least two alternative, future-state models against your baseline model. 7. Step #7—Compare the as-is process with two alternative state processes and two alternative state processes against each other: Investigate the performance of the redesigned process in terms of the cycle time, daily throughput and resource utilization of key resources, such as the x-ray rooms and the dark room(s). Also look at the resource and activity utilization statistics and queue statistics with 95 percent confidence intervals as before. What are your conclusions? Is the new design significantly better than the old process with regards to the cycle time and throughput? Are any drawbacks obvious? Helpful Hint: Although regression models or ANOVA models would be most appropriate, the SimQuick book (pages 98-106) walks you through statistical modeling with descriptive statistics and the T-test for unequal variances. Develop the appropriate T-test comparisons and report. Remember, to fully assess the likely effectiveness ofalternative, future-state models vis-à-vis a baseline model or two or more alternative models, we need to compare the results using statistical inference methods to provide us assessments of process averages as well as likely variation in our processes. You need to report on both process averages and process variation. PART III: REPORT TO MANAGEMENT You are the leader of the process improvement team, and you were chosen to write-up the results from the process simulation and report your findings and recommendations to the management team of the hospital. In particular, the CEO, the CFO and the CMO are members of the executive team and have their concerns you must address. The CFO is concerned with resource utilization and throughput rate and is skeptical about the need to hire additional staff while the CMO is more concerned with patient satisfaction with the X-ray process. You need to address both types of concerns as well as support your findings with analyses drawn from your process simulations. Your grade is assessed based on your report to management and whether you followed the steps in the simulation process. Remember, the CEO, CFO, and CMO do not want the technical details about how you ran the analyses; they’re not the technicians and really don’t have any interest in the technical aspects of process simulation models. They primarily want to know about the current process and whether your recommendations save time, save money, save resources or create potential new revenue. Don’t bore them with the details. Save the details for a technical appendix, footnotes, or end notes.Table-1 Sample Arrival Times of Patients1 Patient # Time of Arrival (in min. from time 0) Patient # Time of Arrival (in min. from time 0) 1 6.30 31 197.89 2 10.13 32 205.50 3 17.07 33 215.42 4 17.09 34 219.95 5 23.94 35 223.50 6 26.06 36 233.33 7 27.65 37 234.89 8 29.21 38 239.20 9 41.65 39 244.29 10 44.69 40 247.29 11 60.07 41 249.90 12 70.34 42 250.25 13 70.73 43 256.34 14 74.32 44 257.90 15 84.59 45 268.97 16 91.77 46 276.82 17 95.78 47 280.43 18 98.20 48 281.94 19 117.24 49 293.23 20 122.85 50 299.79 21 130.58 51 303.75 22 137.46 52 306.58 23 139.76 53 308.13 24 142.52 54 314.06 25 150.70 55 322.82 26 151.95 56 326.51 27 154.74 57 338.21 28 157.48 58 339.91 29 193.25 59 353.65 30 195.46 60 359.79 1 Note: These are arrival times, not interarrival times. Interarrival means the time between arrivals, so you need to subtract the time between subsequent arrivals. For instance, patient number 29 arrived 193.25 minutes after the clinic opened while patient number 30 arrived 195.46 minutes after the clinic opened. The interarrival time between patient number 29 and patient number 30 is =195.46 – 193.25 or 2.21 minutes. From the interarrival times, you should run the descriptive statistics for interarrival times in MS Excel and note the mean and standard deviation from the statistics. You should also create a histogram in MS Excel to try to determine which statistical distribution best fits the data. At a very minimum, the mean or average interarrival time and the appropriate statistical distribution are needed to estimate the interarrival time when patients enter into the clinic.Table 2 Activities in the Current X-Ray Process2 Activity Description Type 1 Patient leaves physician’s office with instructions and enter the X-ray center at an unspecified interarrival rate. Start of X-ray process 2 Patient takes a number and waits for the front desk assistant to call. Lobby (Waiting Line) 3 The front desk assistant fills out a standard form based on information supplied by the physician and the patient. The patient then leaves the front desk, and queues up in front of the X-ray labs. Work Station (Value-Added) 4 The patient enters the X-ray lab, undresses, and an X-ray technician takes the required X-rays (all done in the X-ray labs). Work Station (Value-Added) 5 A darkroom technician develops the X-rays. (Assume that the patient accompanies his/her X-rays). Work Station (Value-Added) 6 An X-ray technician checks the X-rays for clarity. (Assume that the patient accompanies his/her X-rays.) Work Station (Inspection) 7 If X-rays are not clear, the patient returns to the waiting room in anticipation of repeating steps 4, 5 and 6. Historically, the probability of rejecting X-rays has been 25%. If the X-rays are acceptable, the patient proceeds to activity 8 and the X-rays are put in the outbox, where eventually a messenger service will pick them up. Decision Point 8 Patient waits for the front-desk assistant to out-process. Buffer (Waiting Line) 9 The patient out-processes with the front-desk assistant and goes home. Work Station (Out-Process) 10 Completed X-rays are taken to requesting physicians’ offices. Finish Process 2 You may need more elements than listed here for the SimQuick model to work properly. You need to assume that the patient follows the x-ray, meaning that the patient waits as the x-ray waits to be processed. In short, you will need more buffers than indicated by table-2.Table 3 Resource Data for X-Ray Process3 Resources Activities # of Units Available X-ray technician 4 and 6 3 X-ray lab 4 2 Darkroom technician 5 1 Darkroom 5 1 Front-desk assistant 3 and 9 1 Table 4 Standard Times for X-Ray Process Activity Activity Time Distribution Parameter Values (in minutes) 1 See Table-1 See Table-1 2 Non Applicable Non Applicable 3 Uniform Max=5, Min=2 4 Normal Mean=15, SD=5 5 Normal Mean=10, SD=5 6 Normal Mean=3, SD=1.5 7 Not Applicable Not Applicable 8 Not Applicable Not Applicable 9 Uniform Max=5, Min=2 10 Batch Process Daily 3 Here’s a word of advice: keep the resource modeling as simple as possible. Having a resource listed in the table doesn’t necessarily mean that you have to use the resource modeling capabilities in SimQuick to model resources. For instance, if you have two of a certain resource available and two workstations that need the resource, it’s generally reasonable to assume that each workstation would have a dedicated resource. Resources don’t necessarily need to be shared. If, however, you either want to or if the situation requires that two or more workstations share a common resource, such as using crosstrained workers in exercise #10 in the SimQuick book, you need to use the resources table(s) in SimQuick to model shared resources.