FILE NAME: FLOW2
FUNCTION: Scheduling 2 work stations in series
OVERVIEW: FLOW2 computes a schedule for a group of jobs to be processed through two work stations in series. Each job must be processed in the order station 1, station 2. The usual objective in this type of scheduling is to minimize makespan, defined as the total time from the beginning of the schedule until the last job is complete. If there are n jobs, the number of possible schedules is n!. With only 10 jobs, there are 3,628,800 possible schedules.
FLOW2 uses a priority index in column F to sort the jobs. The index is computed as follows for each job: If the time on station 1 is less than time on station 2, the index is the time on station 1. Otherwise, the index is 1000 - time on station 2. When <CNTL> C is invoked, the jobs are sorted by index number in ascending order. This procedure gives exactly the same result as a widely-used manual method (Johnson's rule):
1. List the jobs in a table.
2. Find the smallest processing time in the table.
3. If the smallest time is for the first station, place the job in the first available position in the sequence. Otherwise, place the job in the last available position. Ties are broken arbitrarily.
4. Remove the assigned job from the table and go to step 1.
Although the mathematical justification is beyond our purposes here, the scheduling rules make sense intuitively. We always put the job with the lowest station 1 processing time first. This allows station 2 to go to work as soon as possible. Furthermore, we always put the job lowest station 2 processing time last. The reason is that total processing, for all jobs, can be finished as early as possible after the last operation is complete on station 1. In other words, it is best to have both stations working at the same time as much as possible.
INPUT: List job names in B12..B16, processing times on station 1 in C12..C16, and times for station 2 in E12..E16. Pressing <CNTL> A sorts by job name, <CNTL> B by random, and <CNTL> C for the optimal sequences.
OUTPUT: Beginning and ending hour numbers for each job are shown in C21..F25. The Gantt chart at the bottom of the worksheet converts the hour numbers to times of day. For each sort, makespan and average flow time are reported in cells I5 and I8.2. WORK-STATION PRODUCTION MODELS:(cont)
MODIFYING THE WORKSHEET:
To add jobs, insert new rows at the row for job C in each of four sections of the worksheet: rows12-16, 21-26, 33-37, and 39-3. In each section, copy the row of formulas above the inserted rows through the last row in that section. For the first section, just copy the formulas under G (Pri) and I (Random number). The other columns for job and processing times need to be input for each new job. You can extend the Gantt chart to show additioNal work days by copying range 030..044 to the right. To modify the Gantt chart for different work schedules, edit row 30 to show a sequence of work hour numbers, interrupted by labels for lunch and other breaks, In row 31 enter a sequence of times corresponding to the work hour numbers and labels.
The "actual" Gantt chart that starts below is an example of copying over the planned Gantt chart and entering the actual start and stop times of events -- thus providing a key report for variance.
1. Get Started
2. Workstation Production Models
3. Scheduling 2 Workstation in Series
4. Scheduling 3 Workstation in Series
5. Scheduling 4 and more Workstation in Series
6. Scheduling 2 Workstation in Parallel
7. Project Management - Daily Schedule
10. Gantt Charting