Seeking advice regarding missing and cross threaded bolts

OK guys and gals, here is the problem. I work in the automotive industry and our biggest RPN issues are missing/cross threaded bolts, we have tried several things to combat this issue all have met wih moderate success at best.
1. Training (have always done) - All people putting in bolts have been trained, retrained, retrained, and retrained.
2. Audits (have always done)- Audits are performed to make sure that the associates are following the correct sequence of their operation manuals.
4. Click wrench (have always done) - After bolt is tightened, ratchet clicker wrench is used to force operator to check for bolt presence and torque, hopefully a cross threaded condition is caught.
5. Sequential verification (have always done) - An associate other that the one that put in the bolts and used the ratchet clicker puts a verification mark on the bolts to confirm presence of bolt and hopefully detect a cross hreaded condition)
6. Atlas copco electric guns (used on several trial applications but discontinued) - Worked well to catch both cross threaded (bolt angle) and missing (limit switch interlock) bolts, problems were too slow, durability, poor ergonomics for our application, expensive.
7. Yakota pulse air guns (currently used for all bolt applications) - Work well to catch missing bolts (limit switch interlock) problems include cannot be programmed to catch a cross thread condition in our application because the torque angle does not require enough countable pulses.

Yet with all of these human and mechanical controls in place, we still miss an occasional bolt but mostly cross thread them.

IDEAS APPRECIATED

[Jim Biz]

Hmm - just "food for thought" certainly don't want to compound the problem with questions you may have already addressed.
BUT

Seems as I read this that you have spent much time & effort to control the assembly methods and end results during and -after the fact... but have yet to discover the primary root cause?

Why are they missed? Why/how can they cross thread in the first place? Assembly Cycle speed? - miss-formed or mis-enginerered thread form - on the bolts themselves the tapped hole (or mating nut)? - possible a "finer thread" would help? Possibly an increased chamfer added to the thread lead?

Are the bolts that are being missed/crossed in random or consistant locations that are not out in the open and difficult to access - easily overlooked? Are piecwork pricing - or daily quantity requirements - part of the equation? OR (and this may be way out in left field) would it be possible to attach what you have with a completley different approach? - rivets - other means?

Regards
Jim

Jim: to answer your questions
Root cause: Human error
Bolts: Some flat tipped, some with a pilot. All nuts are the old fashioned kind of weld nuts. Sadly, we cannot alter the design, as it was designed in Japan (As you are aware, they dont make mistakes and there is no better way than theirs. Just ask them) - We have tried to switch all nuts/bolts to an extruded nut and a self tapping bolt but the design section had several reasons why this was an inferior design (in my opinion, the inferior part was that they supply the nuts and bolts but with extruded nuts, there is no nut and the self tapping bolt is under patent to a US company so they could not make it)
Missing/cross threaded bolts - In various product lines done by various operators. Sometimes visible on the complete assembly, sometimes not.

Perhaps you can see our frustration

[Jim Biz]

Surley can see the frustration - but somehow can't buy-in on Human Error as root cause for for an ongoing or re-occuring problem.

Possible to "track" - or gather data on which ones are crossed/missed most often? - investigate largest contributing areas (do a pareto) - - initiate additional training or do a batch/lot "pre assembly screening" of the bolt&nut fit-up? - or a team improvement project there first?

Regards
Jim

[Sam]

When you seem to come to a dead-end and the thoughts that you and your mates have just don't seem to work; Try "Benchmarking". Very effective, look at companies with similar processes; how does the "big 3" install the tire & rim assembly, how does the pit mechanic change a tire so quickly?
Have you tried DOE. Excellent tool.
Have you asked your Japanese counterpart for assisstance? After-all they did perfect the art of "Problem Solving".

[Rick Goodson]

In a past life I had a similar situation. We were building on roller line and had to torque nine bolts. If memory services there were three different torque requirements. We used a physical stop in the roller that would not let the product continue until it dropped. The stop was connected to a computer which also had the torque wrenches connected to it. Software control would not drop the stop until all torques were registered.

[CarolX]

pd - a couple of thoughts

Don't get too discouraged - remember, we are human after all, and as long as there is a human factor, mistakes will be made (although I know this perspective doesn't always satisfy the boss or the customer).

Is there a way to fixture the part that will allow the air gun to detect a cross thread?

Good Luck,
Carol