Acceptable Error Rate?

[New_B]

I manage a very small R&D department, developing materials testing products. I am frankly CAD illiterate. I am in this position as a result of my testing and product development background. We have one person who creates and maintains most of our drawings and there is concern about his competency. There seem to be many issues about his drawings that cost us money in lost time and unacceptable parts from vendors. We now have someone with CAD experience checking his drawings closely and it seems better at this point.

 

I am under heavy pressure from our CEO to have his performance improve or get rid of him. The problem is, I do not have any idea what an acceptable error rate is. I don’t expect drawings to be 100% error free, but how much is too much? I want to be fair, but I also don’t want one person to torpedo our department. Any wisdom to share?

[Ryder76]

Acceptable error rates vary. One place I worked our department had less than .5% error rate, but we were constantly told our department was lousy and cost project money. Not true because 99.5% of the bloody markups and reason for return was the engineer was re-inventing the wheel more than once. They didn't seem to understand that drafters don't generate the work - engineers do. This place also did not have an official 'checker' - usually an engineer that knows drafting as well as the discipline.

 

Is anyone checking the work that comes out of the drafting department? The drawing should always be checked by the drafter and then checked again (internal to the department) before being sent to an engineering 'checker'. If it is a small company and there is no one to check then the error rate will be higher going to the end user. This usually looks bad even if it is understandable due to a lack of a checking process.

[ReMark]

Who was responsible for checking this CAD drafter's work? Is he considered a CAD draftsman or a designer. Each has different levels of responsiblity.

[New_B]

Ryder 76:The person checking his drawings started in a peer position, although he has a higher degree (Bachelor's in ME). He is now our production manager, but still checks the drawings. I have looked at the drawings in the past, but clearly, I'm not equipped to spot thing like tolerance deviations, improper finish call-outs, and even some dimensional issues.

[New_B]

ReMark: The subject functions as both designer and draftsman. He basically is given a set of requirements by me or through me to create a device to test materials to an ASTM standard.

He seems to do ok in most stages until it comes to the finished drawings that go out the door to fabricators. When the parts start coming in wrong, the blame game starts. They might say, "we didn't deburr that part because it wasn't stated on the drawings." or "the right angle isn't 90 degrees because it isn't specified." For these things, I have to believe they know better and are looking for excuses. However, there have been a number of instances where dimension and tolerance errors were definitely in the drawings and the fault was ours. In an operation as small as ours these errors are very costly, and we end up eating some that are probably the vendors as well.

We are using the latest version of Solidworks and I have wondered if he isn't in over his head with the complexity of the software. He has had numerous training courses, but I wonder if we would be better off with a simpler software package such as Alibri.

[ReMark]

It's difficult enough in this economy to keep a business going without adding to the problem by self-inflicting needless errors.

 

Is there an form of error checking in Solidworks?

 

I hear what you're saying about "looking for excuses". Sometimes I think this is just the de facto "norm" for the way some businesses operate. Blame everyone else and hope to get a better price.

[New_B]

ReMark: There is a function in SolidWorks that tracks things like hole dimensions and tolerances, but it seems that it is complex enough that not even SolidWorks Tech support agrees on how it should be used. Our guy decided to start using it, and instead of testing it on a dummy or obsolete set of drawings, he fouled up a current set of working drawings just as we were ready to send them out for quotes. That's why I wonder if we need to use something simple. In the course of the ass-chewing I gave him, I said we would go back to drafting vellum and Koh-I-Noor pens if we had to. He seemed shocked that we could turn our back on new technology, since "new" is obviously better.

[Cad64]

The program is just a tool, and it's only as good as the operator controlling it. I really don't think switching to a simpler program is the answer. I would suggest switching to a more competent operator. You stated that this guy has taken numerous training courses but is still unable to produce quality drawings? Then I would say it's time to start interviewing for a replacement. You don't necessarily have to fire this guy, just hire in a more experienced person and let this other guy train under him, depending on your cash flow situation that is. Maybe this incompetent guy will need to take a demotion and a pay cut, but hey, at least he will still have a job.

[ReMark]

Maybe a different approach is in order.

 

You say the CAD operator has participated in a number of training sessions. Perhaps he does not do well in a group environment and would benefit from some one-on-one time with a recognized expert. You know what the problem is. It's stupid errors. And you know the source of the errors. It's the operator. You must have a local authorized Solidworks rep you can call. Ask him for a quote for a half-day's worth of side-by-side training with your man. Tell him the types of problems you're encountering. He can gear his training session towards eliminating or at least minimizing the errors. Finally, if management approves this plan, sit the operator down and tell him "This is your last chance buddy. Your (his) errors are costing the company money (maybe even lost business) not to mention headaches." Inform him of the new plan and then cut to the chase..."Shape up or ship out. You'll (he) be reviewed in 30 days and a decision will be made too retain or to dismiss." Now everyone knows what is at stake. There are way to many people out there right now that would only be too happy to step in and take over where he left off.

[DANIEL]

A standard checking practice would help you greatly, not just in reducing errors but in seeing who's not getting the job done satisfactory. Be prepared for a learning curve however, the first few checks are likely to turn up a fair bit more errors than once they have been acclimated to the process. Heres what I would suggest.

 

1. drawing is drawn by person A

2. drawing is checked by person B

3. drawing is corrected by person A

4. drawing is checked by person B

5. repeat step 3 if nessicary

6. drawing is check by person C

7. repeat step 3 if nessicary

 

It may seem a little excessive but it will eliminate nearly all your errors, yes, nearly . Once they get use to the process it won't seem nearly as taxing either.

 

I would also recommend rotating the checking duties, allowing those that do the drawings to look at things from the other side, it will improve their drafting most likely.

 

But most importantly, if someone needs to be let go, this process will give you a good idea of who it needs to be.

[New_B]

Thanks all, for your input. I can't say I found a "magic bullet" answer, but I'm a lot further along with a better perspective than before. This guy is talented in other areas, so I hate to just kick him to the curb, but we definitely need to have a different system for producing drawings.

[ReMark]

Good luck. Let us know how things turn out if you don't mind.