Some packages that pass through a primary sorter cannot be routed to a secondary sorter for a variety of reasons and must be manually processed. Suppose such packages are routed to a circular conveyor as shown in Figure 16-4. Packages proceed around the conveyor to a workstation. There is no package waiting area or buffer at a workstation. If a package arrives to a station that is processing another package, it stays on the conveyor to the next station. If the package is not processed by the last station, it recirculates to the first station.
The purpose of the simulation study is to specify the parameters of the manual system to minimize package lead time. There may be either 1, 2, 3, or 4 workstatons employed. In addition, waiting areas for up to three packages may be placed in any fashion among the workstations. Cost considerations make more buffer spaces and fewer workstations the preferred design. Determine the number of workstations, the number of buffer spaces, and the location of the buffer spaces.
Relevant information is as follows:
Time between package arrivals: Exponentially distributed with mean 1.6 minutes. Package processing time: Exponentially distributed with mean 4.0 minutes.
Conveyor Segments (Assuming a Four Workstation Configuration)
Assume that conveyor speed is 0.25 feet / second and that packages are 2 feet in length. The time period of interest is 40 hours.
Case Problem Issues
1. Count the number of possible alternatives. Is it reasonable to simulate all of these?
2. Which alternatives should be simulated to make sure the best or at least a good alternative is identified?
3. What performance measures in addition to package lead time are of interest?
4. What operating rules could be added to the system to guard against excessive lead times for individual packages?
5. What is the minimum number of workstations required by the system?
6. Discuss how verification and validation evidence can be obtained.
7. What is the purpose of having buffer space in front of workstation? 8. How is the arrival of a package to a workstation modeled if:
a. There is no buffer space at the workstation.
b. There is at least one buffer space at the workstation.
9. What is your initial guess as to the best placement for the buffer spaces? Does the simulation study confirm your guess?
10. Tell how to compute the lead time for a package as a function of the number times it travels completely around the conveyor within the simulation.
11. What is the radius of the conveyor: radius = circumference / 2 p ?