Consider a manufacturing process that involves drilling holes in car engines. The manufacturing system operates from 8 a.m. until 11 p.m. in two shifts (you can ignore the turnaround time between the shifts). Due to a recent increase in the demand for the parts, the managers are considering replacing the current equipment and have hired you to evaluate the following two options:

1. A drilling machine with a capacity of two (i.e., two engines can be processed at a time). The processing time is uniformly distributed between 1.6 and 3.2 minutes.

2. A fast drilling machine that has a capacity of one but the processing time is uniformly distributed between 0.8 and 1.6 minutes.

Your task is to evaluate the performance of the two machines and make a final recommendation as to which machine to purchase. Assume top management is interested in two performance metrics: the throughput (i.e. number of parts produced per day), and the amount of work in process (i.e. average inventory in the system). Engines arrive at a rate of 40 per hour and inter-arrival times are exponentially distributed. Part transportation times are negligible.

Assignments:

1. Watch the following two videos (approximately 30 and 20 minutes, respectively):

a) http://simulationandsimio.org/node/31

b) http://simulationandsimio.org/node/32

2. Build one Simio model for EACH machine.

3. Run each model for 50 replications, use a warm-up period of two hours, and make a recommendation based on your results.

4. For both machines, run several experiments varying the number of replications (i.e. 50, 100, 150, ). For each experiment, report the 95% confidence interval of the two performance metrics of interest. You can build a table or a plot (in Excel) to report your results. What number of replications would you recommend to use for this problem? Why?