Looking for ideas:bigwave:
I am using our Business system to capture history on inspection hours charged to part families we manufacture.
What information I have;
1. # of orders within part family. ex. 50
2. # of part numbers related to part families 10
3. Total number of parts ordered. ex. 500
4. Total of inspection hours charged to part family. ex 22.50 hrs

All data was captured for a 12 month period.
I am doing a Pareto chart with several of these part families. The end result will be to pareto, by part family, the most cost effective approach to implement in-process inspection activities.
I have my theory, but would like to get some input. Keep in mind that some families may have 2 parts with 20 hours, and some have 20 parts with 30 hours. The more parts the longer the Classification of Characteristic process will be.
Thanks,
Julie

Hi Julie,

As it is unclear to me what differences there are in your inspections' complexities, adequacy of available tools and reference materials, and competency of inspectors, it's hard for me to recommend an approach except:

Reduce variation. If one inspection step runs more smoothly than the next one, ask why, and how could improvements be made to the problematic one? If inspectors' capabilities vary, would the organization be served to equalize all their skills, or to place the more skilled inspector sin complex and demanding roles? These are cost saving inspection strategies as I know them.

When we better streamline all of the (man, material, machine, environment) aspects of inspection, we can more accurately analyze how much time each inspection step requires for cost estimating. Until the variation is removed your etimates will be guesswork.

Have you eliminated variation in your inspection yet?

P.S. We use PeopleSoft to track costs.

Hi, Julie.

One thing I would warn about with this type of data is that sometimes this data comes from managerial accounting systems. Often, what happens is that a manager must allocate their costs to product A, B, Z, whatever. This allocation can range from a very accurate accounting, to a good (or not so good) guess, to just making the numbers work. If the numbers are reported from the floor, there will likely be varying degrees of accuracy, as well.

So, if you have not done so already, please take a few minutes to (tactfully) determine the quality of the data you are working with. It would be bad to spend all of this time only to come up with something that has no resemblance to reality.

My teams' task is to prioritize (pareto) which part families to start the Classification of Characteristic process with.

I am looking into data to support this process. Such as the inspection hours. I do not have inspection variation information, except for the business system entries.

We are going to an In process inspection environment so developing the C of C's is the next step for our team.

Thanks Craig!
I have found exactly that problem. I am only as good as my data in this project. I have spent many hours preparing the information and interviewing operators, inspectors and their supervisors. I think what I have now is as accurate as it can be.
The team likes all the data I have, but they are unsure of which way to utilize the information.

Thanks Craig!
I have found exactly that problem. I am only as good as my data in this project. I have spent many hours preparing the information and interviewing operators, inspectors and their supervisors. I think what I have now is as accurate as it can be.
The team likes all the data I have, but they are unsure of which way to utilize the information.

Smart. Very smart.
:agree1: :applause:

Looking for ideas:bigwave:
I am using our Business system to capture history on inspection hours charged to part families we manufacture.
What information I have;
1. # of orders within part family. ex. 50
2. # of part numbers related to part families 10
3. Total number of parts ordered. ex. 500
4. Total of inspection hours charged to part family. ex 22.50 hrs

All data was captured for a 12 month period.
I am doing a Pareto chart with several of these part families. The end result will be to Pareto, by part family, the most cost effective approach to implement in-process inspection activities.
I have my theory, but would like to get some input. Keep in mind that some families may have 2 parts with 20 hours, and some have 20 parts with 30 hours. The more parts the longer the Classification of Characteristic process will be.
Thanks,
Julie

OK, Julie, it sounds like you have a lot of parts to look at. Is there one group that dominates either the number of parts produced, the number (or number of hours of) inspection, or the most rejects? It might be a good idea to pick a few to start with instead of trying to do all at once.

Of course, it you already have all of the data in a spreadsheet or stat package, the analysis time may be so short that it is not a consideration. So, what form is the data in now? How much work/time to "plug and chug" all of it?