Some companies want to prepare Weibull plot of their production output each month. Looking at the process each month is similar to getting a new “sample” of what’s happening and this foretells how the financial results will go into the record book.
The key issue is to identify undesired patterns of substandard performance and make timely corrections. Each Weibull production plot provides a visual scorecard of how you’re doing and the line slopes, segments, and cusps on each Weibull production plot tell about “mixed failure modes” from a variety of cause and effect problems that take away from consistent output. You want your production output to be very consistent (steep b’s) with no cusps, and large output (high h’s). What you get is often far different than what you want and this provides you opportunities for improvement.
When you have more than five processes to study, this can be a tedious task of making the production probability plots. WinSMITH Weibull (WSW) probability software can help automate the task (by use of batch capability) to quickly perform the task in the background and make presentation quality graphics for import into PowerPoint. This procedure works with software version 3.0T and above. If you’re interested in making “Board Of Directors” graphics, then you need to read about the HGL graphic conversion details for PowerPoint.
Two files are listed below for download that will help you automate your program so you can test drive the results.
The problem is how do we automate our analysis effort to produce the four Weibull production plots as shown in Figure 1. Here are my times for accomplishing the analysis and getting to the point of having Figure 1:
· Copy four production datasets to the Excel spreadsheet JUN98PRB.XLS = 8 minutes.
· Run the macro in Excel to save each Excel dataset as a *.dat (PRN) file, and then call WSW to analyze the data and put the probability files into *.HGL format = 1 minutes
· Import the four *.HGL files into PowerPoint = 3 minutes (Using ImageStream graphic filter—for Office 2000, this filter will not work!! You must use HiJaak Pro version 5.0+ or Correl Draw version 4.0—not version 6.0)
· Convert the graphics into presentation graphics = 60 minutes or 80% of total time expended—when graphics go to top management, they’d better look good because they are your sales tools and you’ll only get a small amount of time to sell your ideas so you REALLY need them to look good! (Fulton Findings has improved this portion of the batch processing for great looking, presentation quality graphic output.)
· Convert the PowerPoint file to JUN98F1.GIF = 1 minute
· Publish Figure 1 from PowerPoint on a web page for posting at a later date via FTP = 3 minutes
All of the plots have different scales. All four plots are shown on a single plot in Figure 2 along with the sum of the four outputs. The line showing the sum of the results displays the typical manufacturing problem—over promised output and under delivered production. If you naively believe you will find the total output by summing all the demonstrated outputs you get (2000+1000+200+100) = 3300 K-Lbs/day will exceed the real combined output of ~3,000 K-lbs/day as shown in the graphic for the sum of 1-4—thus you’ll come up short by ~10%!!!! Tomorrow will never come in which you can recover your under production!
Each plant output shows problems by inspection.
· Plant 3 (the next to the lowest output) is the plant for experimentation—if you make a change, you’ll be able to quickly see the results.
· Plant 2 (the biggest output plant) is unruly! The efficiency and utilization losses are tremendous and the hidden factory is very large.
· Plant 1 has a severe reliability problem which begins at 30% with substantial cutback problems.
· Plant 4 (the smallest) has a reliability problem at ~55%, a utilization and efficiency problem, and severe cutback problems.
The amount of information you obtain from Weibull production plots is large and the effort to make the plots is not too great. Without too much effort, you can produce the plots on a monthly/quarterly basis for a communication tool to persuade people to make improvements.
You need two files to make this batch operation work successfully. The JUN98PRB.XLS and AUTO.BAT files are ZIPPED into a single self-extracting file called JUN98PRB.EXE which is 310K in size for download. Double click on the ZIP file using Windows Explorer and when the file unfolds into the C:\SMITHW directory, the XLS file will be ~1 Meg in size and the AUTO.BAT file is very small. You can repeat the analysis process shown above.
Other pages you may want to visit concerning similar issue are:
On Process Reliability As PDF Files For No-charge Downloads
- New Reliability Tool for the Millennium: Weibull Analysis of Production Data
- Process Reliability and Six-Sigma
- Process Reliability Concepts
Batch file processing of production data can save time. The new batch file processing capability is a new feature for WinSMITH Weibull and it will soon be enhanced for process reliability problems to reduce the time required for presentation graphics.
Refer to the caveats on the Problem Of The Month Page about the limitations of the following solution. Maybe you have a better idea on how to solve the problem. Maybe you find where I've screwed-up the solution and you can point out my errors as you check my calculations. E-mail your comments, criticism, and corrections to: Paul Barringer by clicking here.
Technical tools are only interesting toys for engineers until results are converted into a business solution involving money and time. Complete your analysis with a bottom line which converts $'s and time so you have answers that will interest your management team!