Determine if your final inspection practices are 'waste' or not

[D.Scott]

Ah, so, Toyota must use the same philosophy that we do (we just never called it anything other than the right way to do things) The point that I am trying to make, and maybe I am too subtle, is that everything has a final inspection somewhere. It might be on the line, during inspection, but somebody has to say, yep, that's ok, it can go on to the next process. I am totally against taking something, putting in queue for shipment and then pulling it out just to measure it to make sure it is (still) ok.

No arguement from me.

Dave

 

[duecesevenOS - 2009]

More likely they bought from a 1st tier supplier who inspected all kinds of quality into it. You don't believe they actually made it themselves do you?

As far as "quality built into the process" you're just wrong. Perfection wasn't built into the process - true. To state that there isn't any quality built into the process because a defect was found during inspection, is an insult to every quality professional that works in any industry that has ever produced a defect. That would be just about...everyone! :mg:
It would be better to state that there isn't enough quality built into the system, that it is subject like all things, to improvement.

Your right. I guess what I should say is that quality was not built into the process to prevent that particular defect. There is always going to be knew defects.

Toyota's 1st tier suppliers are mirror images of Toyota itself. Toyota spends as much time working to make their suppliers better as they spend making themselves better. Your right, they don't make the steering wheel, but if a steering wheel was shown to have a defect at Toyota, I guarantee they wouldn't start inspecting the steering wheels as they came in. They would send their own engineers to the supplier and make the supplier better.

 

[SteelMaiden]

Your right, they don't make the steering wheel, but if a steering wheel was shown to have a defect at Toyota, I guarantee they wouldn't start inspecting the steering wheels as they came in. They would send their own engineers to the supplier and make the supplier better.

So, maybe the steering wheel is not the right example, but still, that steering wheel production process (where ever that physical process happened) would have included some sort of inspection. Just because Toyota didn't do a final inspection on it, they sure as goshdern expected somebody to make sure it was right before it was sent on.

you also have to remember that some materials are inherently going to have defects, it is whether or not the defects are within the allowable tolerances that count. Those processes rely more heavily on inspection than the company that only assembles all the other things together. The closer you get to raw material, the more "imperfect" that product is. Creating a process to ensure quality in those cases relies heavily on in-process inspection/validation.

 

[duecesevenOS - 2009]

So, maybe the steering wheel is not the right example, but still, that steering wheel production process (where ever that physical process happened) would have included some sort of inspection. Just because Toyota didn't do a final inspection on it, they sure as goshdern expected somebody to make sure it was right before it was sent on.

you also have to remember that some materials are inherently going to have defects, it is whether or not the defects are within the allowable tolerances that count. Those processes rely more heavily on inspection than the company that only assembles all the other things together. The closer you get to raw material, the more "imperfect" that product is. Creating a process to ensure quality in those cases relies heavily on in-process inspection/validation.

Your right of course. At some point in time in the value stream, quality was inspected into the process. It probably wasn't the steering wheel at Toyota, I kind of doubt it was the tier 1 supplier of the steering wheel, but it was probably the materials that went into the wheel. Toyota might be far enough along in their production development that the tier 2 suppliers are not inspecting quality in. I don't know.

Personally, the facility I work at does a final inspection that finds defects way too often. We are still building quality into our own process so we don't look to our suppliers too often.

The point is: Our goal is to get rid of the necessity for that final inspection. And then get rid of the inspection needed earlier in the process that prevented that necessity....and so on, and so forth. Eventually, your facility won't be doing any inspection because your suppliers are doing it.

That could be a huge understatement: EVENTUALLY. Toyota has been doing this for over 50 years and they are GOOD. The lean consultants are always hyping these ultra fast turnarounds that come from picking the low hanging fruit but a real implementation of TPS with the type of benefits that Toyota has is looking at least 10 years into the future.

The idea of looking at an ideal state without final inspection isn't so that you will do the right thing for today or tommorow...It's so that you will do the right thing for 10 years from now.

 

[SteelMaiden]

Eventually, your facility won't be doing any inspection because your suppliers are doing it.

Won't ever happen, cuz I am one of the suppliers that everyone else is expecting to do all the inspection. There is nobody lower than I am on the totem pole and you sure the heck wouldn't like it if we quit performing chem analysis, physical property testing, etc. etc.

Automotive is a strange creature, and very little "made from scratch" production happens in the automotive world. Most things are purchased and put together. You are putting together a sandwich where all the ingredients are there, you just have to stack them in the right order. I am the one who has to make the proper mixture of eggs and oil, salt and other spices to make the mayo. or the mustard, or the ketchup, or the pickles. You definitely want me to inspect my work.

 

[Jim Wynne]

Automotive is a strange creature, and very little "made from scratch" production happens in the automotive world.

I think there are a lot of people who work in metal stamping, sheet metal, machining, injection molding, etc. who would disagree with that statement. Remember, the components you use have to come from somewhere.

 

[SteelMaiden]

I think there are a lot of people who work in metal stamping, sheet metal, machining, injection molding, etc. who would disagree with that statement. Remember, the components you use have to come from somewhere.

I used to work for a supplier to automotive. I meant that the suppliers send things to the auto maker, and they are assembled there. I seriously doubt that you'll see an automaker producing the sheet metal, machining the sheetmetal into parts....as you just said, that is done by the metal stamping, sheet metal producers, machining shops, injection molding companies. Believe me I know the components came from somewhere but I know that GM, Ford, Toyota et. al. did not manufacture the steel rod that was used to produce the gear shift lever.

 

[Bev D]

so to the OP's original question: "How do we counter the argument that the final inspection process is unnecessary and wasteful?"

Well by definition in Lean final inspection (wherever it occurs) is always waste. it adds no value to the product. so there is no argument to this operational definition. HOWEVER, there is a solid argument to the unnecessary description. if defects are present at the final inspection location it is necessarry to detect and remove, rework or repair them so they dont' get to the Customer. Having final inspection when defects are presetn is less wasteful than not having it. That's your argument.

Of course the goal of lean is to always be driving towards the ideal state. so your lean consultant or leader should be driving you to eliminate the NEED for final inspection. No lean leader who is not a hack would insist on reducing the defects. If all they say is "get rid of final inspection" they are hacks. (By the way any customer who insists on the automotive PPAP process for their suppliers - or the equivelent - is NOT practicing Lean.)


Lean drives towrds the ideal state but even Toyota has never gotten there. It's the journey. Reduce the amount of waste: 5S audits are wasteful - but less wasteful than not having a 5S environment. and the goal is to make 5S so ingrained that the audits are not necessarry. At Honda we didn't do 5S 'audits' but the requirement was so ingrained that anyone who walked by and saw even a gum wrapper on the floor would pick it up and throw it in the trash barrel. those around would be embarrassed - it was a rare event. TPM is truly preventive: monitoring to detect and then repair wear BEFORE the equipment breaksdown or makes bad parts...so it's monitoring and 'inspection' is far less wasteful than final inspection.

 

[Icy Mountain]

I like the concept of "waste" and "unnecessary" waste. Let us also consider the concepts of "justified capital expenditure" and "rejected capital expenditure". Here we go:

$1 = per piece cost of 100% final inspection (finding 0.5% failures)x 8000 / year
$8,000
$240 = per piece cost of a failure reaching the customer x 8000 / year @ 0.5%
$9,600
This is necessary waste "saving" me $1,600 per year plus 40 failed units in the hands of customers. The circuit board is first assembled by machine (near flawless), then assembled by hand (about half of the failures are generated here). Finally, the board is assembled into a housing that protects it from the environment and also serves as a heatsink.

Now, given $8,000 per year of "waste", this is what the lean hype guys have to work with in order to justify capital expenditures on mistake-proofing. They all have the magic until they realize that any further mistake-proofing is going to take either automation and/or serious design changes to allow for automation. After that, they disappear into the woodwork.

Second lesson:
$4,800 = $60 per hour x 1 hour per board x 5 boards per unit x 16 units per month cost to pre-test circuit boards 1 by 1 to assure 0 defects at final test so that some "lean genius" can eliminate "unnecessary" final test and rework.

$960 = $40 per hour x 1 hour per unit x 16 units per month cost to final test only +
$180 = $60 per hour x 3 hours per defective unit x 1 defective unit per month cost to troubleshoot the 1 defective final unit each month (93.75% yield, how terrible).

Now, that's $4,800 vs. $1,140 to assure that no defective units reach the customer. Would you believe that our "lean" VP was actually going to implement board level test? Only when I pointed out that we still would not save the $960 did he back off this position: 90% of "final test" was actually establishing traceability of the measurements (while making sure that they were correct) so that we could issue the calibration certificates that every customer required.

 

[Helmut Jilling]

I like the concept of "waste" and "unnecessary" waste. Let us also consider the concepts of "justified capital expenditure" and "rejected capital expenditure". Here we go:

$1 = per piece cost of 100% final inspection (finding 0.5% failures)x 8000 / year
$8,000
$240 = per piece cost of a failure reaching the customer x 8000 / year @ 0.5%
$9,600
This is necessary waste "saving" me $1,600 per year plus 40 failed units in the hands of customers.

I agree with your premise. However, I don't agree with the unit numbers as presented.

Does it really cost $1 per piece to do a final inspection? Seems high.

Also, I think the cost of a defect reaching the customer is typically MUCH higher than $240, when you factor in all the time, meetings, responses, and other soft costs that are incurred. If my hunch is correct, the new numbers will reinforce your example significantly.