iSixSigma

Poke Yoke assembly inspection

Six Sigma – iSixSigma Forums Old Forums General Poke Yoke assembly inspection

Viewing 14 posts - 1 through 14 (of 14 total)
  • Author
    Posts
  • #52065

    Grosvenor
    Participant

    Hi guys,
    I am the Lean manager for a small job shop (250 folks) and we have an ongoing problem with quality inspections for our products (oil field support trailers) prior to shipment. Problem being various things-parts missing, or wrong part, or in wrong place-and discovered by the customer. We fabricate almost all of the trailers (weld from raw steel, paint) and assemble, some weldements being outsourced. We have detailed checklists for most of our trailers, but some items still get past. Our volume ranges from 1-2 to 3 per day.Would appreciate an outside the box opinion of a fresh way to attack this problem. Our quality manager does most of these inspections, and he and the department head have to sign off on the final form together. We have another inspection process in the weld shop. Hard to see how these vaiatons get through, but they do.
    I come from a high volume manufacturing background (appliances) so this high product mix low volume process is much different to analyze. Still, these trailers are usually very complex, and it is easy to miss things. Sometimes, the items in question are supplied by the customer and turn out to be the wrong item (which is not our fault, just one more item). I am just asking for your help figuring a new angle of attack, as this quality problem is becoming very serious to our business. Adding another layer of inspectors isn’t a good answer as we are not able to hire any more at this time (although some of our competition has done that).
    Thanks guys
    Chuck

    0
    #182541

    Severino
    Participant

    One way of solving the missing components issue is to utilize pick-to-light systems and barcode your shop traveler (assuming you have one).  As each assembly moves from station to station the operators barcode it in and the p2l tells them what to put into the product.  Provided each workstation is stocked correctly, this could drastically reduce your missing parts.
    What is typically missing?  Big parts?  Small parts?
    How many different configurations of products do you ship?  How different are they?
    Can you change the size of certain items so that you can’t fit them into the wrong location?
    Can you move similar items to a different workstation/area so that they cannot be mixed up by your operators?
    Can you color code your parts so that it is clear what goes where?
    Can you use things like fixed size wrenches at stations in place of variable so that the wrong part cannot be installed on the assembly?
    Can mating parts be designed with a pin and hole so they can only go together one way?
    Its very tough to solve your issue without knowledge of your process, but the above may give you some starting points to consider.

    0
    #182542

    Grosvenor
    Participant

    Thanks for the quick response. I will post a more detailed reply tomorrow, but for now-the trailers go to assembly and stay in one stall to finish. No multiple stations. We have some point of use racks for common fasteners. We are approaching a dedicated space for a couple of our products, but we do not have more than 5 or 6 machines of the same type in succesion. Of those, there are model differences.
    To change sizes etc. involves a drawing change, which has to be approved by the customer.
    There are no typical parts missing. Wide variety.
    This is a job shop, and TPS does not work too well here.(unlike 250 dryers of one model going down an assembly line)  Not enough repetition to set up dedicated work cells, although we have some new products coming in soon. The whole idea is to be flexible to accomodate any type of trailer. We have attacked each problem as it occurs, but there is just too much variation between each trailer to see all of the possible opportunities of error. 
    Will provide more detail tomorrow, and thanks!
    Chuck

    0
    #182598

    Severino
    Participant

    Chuck –
    To say that TPS doesn’t work well at your facility is probably incorrect.  TPS is about reducing waste and every operation has its wastes.  The waste you are most concerned about right now is defects, but the reason you are having the defects is probably tied to a lot of other inefficiencies in your process.
    The best way to solve your issues is to follow these three steps:

    Go See
    Ask Questions
    Be Respectful
    If you watch what your operators are doing you should be able to see problems.  Do they move in an orderly way around the trailer or are they constantly moving from one side of the stall to the other?  Are the parts they need as physically close to the point of use as they can be?  Are their instructions clear?  Are they even looking at them?
    Look at the work area.  Is it free of foreign materials?  Is everything neatly set in order or are things strewn about haphazardly?  Is there equipment or parts there that are never used and only take up space?  Are your machines bolted to the floor or can they be moved around to make the cell match the work?  Without any previous knowledge of the process could you walk into your assembly area and know how material will flow and what activities will be performed where?
    Look at the way work, materials, and information flow into and out of the assembly area.  Is the right stuff coming in?  Is it coming only when it is needed or is it sitting in a queue somewhere?  Is it coming in the correct quantities?
    Focus on the similarities of your items (develop product families).  Sure you may be low volume, high variation but there are likely hundreds of tasks that are repetitive between them.  Focus on them, take the variation out of them so that they are not contributors to your defects.  Then attack the differences.
    Again, you say that TPS doesn’t work well, but it is not about parts moving down a conveyor or about Kanban cards exchanged with a supermarket.  TPS/Lean is about looking at your organization with eyes that are trained to see waste.  Once you truly learn to see and how to take action you will quickly be able to solve a large portion of your problems and probably increase the efficiency of your operation at the same time. 
    Don’t give up, don’t lose hope, and don’t wait till tomorrow to solve a problem that you can solve today.  Cheers!

    0
    #182634

    pedro
    Participant

    Hi Chuck..
    I used to work as quality Manager for a facility assembling Fiber Optics products..  One of the products was a metal sheet cabinet, where we needed to install a big variety of components. We had simmilar problems to the ones you described.
    Our production line, was not set properly,, it was not even a production line,, some kind of batch system  where different people coul be assigned to a particular operation with a poor responsability of the operator on the quality of the product…
    Among a lot of components missing,, we had the labes that should be installed in the cabinet,, a lot of cases with missing labels, incorrect labels..  imaging a cabinet required 24 labels… one person was installing some of them and another on the rest… if he know which ones to install..
    we were really looking for problems with that system..
    we changed to a production line, using pull system, with dedicated people to particular operation, so we could assigned reponsability,, reduced by a lot the amount of product being assembled.
    Also,,,,, we added an operation to kit,, labels… that kit was provided to operator,,, so at the end of the operation the operator should not have any label left in the kit, othere wise a label was missing in the product…. and just to double check.. he needed to count labels to verify qauntity,, and  signed  on a just created check sheet that he performed the operation … this was with the purpose of creating ownership and responsability on quality…  also during weekly quality meeting with employees and inspectors the results were discussed..
    the operation we added to kit labels,, even when it is a not value added operation it was worth it as we stopped the problem,, and reduced,, inspections, reworks,, sorting and complaints…
    Know that every case is different,,,  but I hope this may help.
    Pedro

    0
    #182636

    Alan Harkness
    Participant

    Left field answer – but would it be possible to weigh the item (or the sub-assemblies) prior to dispatch / assembly?  This way, you’d know if something was missing?

    0
    #182637

    Dalziel
    Participant

    My experience here is from the Auto industry – I’ve done mass production, down to early prototype and validation builds. Your situation is obviously closer to the prototype situtation, in terms of static builds, with high complexity and variation.
     
    I would suggest you look to kit parts – identify the common parts, and have a “kit” for them. Doesn’t matter if it’s the size of a tool box, or a shipping container. Just make sure that each section is properly identified – pictures work better if parts are varied.
     
    For “variation” parts, have a second kit. Make sure the build sheet clearly identifies the parts. Again photos are good!
     
    Then as each new job enters the stall, bring the two sets of parts (standard and variation) to it.
     
    Then for inspection, rather than just inspecting the completed job, inspect the kits, get the operators to check they’ve got the right parts for the job. Also, make sure you have easy, but controlled, access to some spares, as parts will always get damaged, or (I’m thinking of small nuts and bolts) lost during assembly.
     
    Final inspection should just be a formality, you need to build the quality into the early stages, before it results in rework, or worse, customer complaints.
     
    good luck,
     
    Dalziel

    0
    #182638

    Severino
    Participant

    I thought of this as well, which is why I asked about the size of the missing components.  It will help with large items, but a missing bolt would go unnoticed by such a system.

    0
    #182639

    Alan Harkness
    Participant

    Yes – an element of scale always applies, but a lot can be resolved by using scales of suitable accuracy and precision.
    What the scales won’t nail, is when the ‘wrong’ part is used – e.g. two left brackets in stead of a left and a right.
    However – worth a thought…!

    0
    #182641

    Fontanilla
    Participant

    I’m assuming you’ve already created Pareto charts of the various failures, correct?  What’s the most common failure?  When does it occur?  Have you also plotted these failures on a diagram of the product and looked for patterns with respect to the location of the failure?  For example, you find a high occurrence on one area as opposed to another.  Have you asked the same questions while looking at individual stalls and individual assemblers?  Are there patterns there?  It’s been my experience that there is always a pattern.  These things are never random.  When they appear random, you haven’t asked the right question of the process.

    0
    #182646

    Ted
    Member

    Chuck:
    I am also from appliance, having worked for a number of the larger presences in North America.  
    I also worked in Barbecue and there as Engineering and Quality Manager I made use of weigh scales at various weights and resolutions to Poka Yoke or at least aid in warning of the issues. 
    They could of course not help if the parts were the same weight or the resolution required was too fine. 
    Basically I used 3 electronic scales of different ranges and sensitivities.
    Low weight, high resolution for the hardware packs, 5-10 Kilos for the accessory packs (grill plates, etc.), and about 100 Kilos for the final product which was boxed but before the lid was put on. The resolution for this was as I recall about 200 grams or .5 lbs.  The scales had to be fairly robust and able to tolerate some rough treatment.
    I do not know if this would work in your industy as you may need to weigh the components and there may not be enough difference among the probable errors.

    0
    #182651

    Grosvenor
    Participant

    Wow, what a great response. I was going to reply the other day, but now glad I waited. Let me try to clarify a bit. I realize now saying TPS “wouldn’t work here” was in fact incorrect. The identification of waste is my primary goal, it’s just that the application of lean methods is very different than high volume-low model mix.  Let me describe a sampling of the defects that have gotten to our customers, as they are not always missing parts but a wide variety of defects. This was a sampling of about 250 over a three year period.:
     Leaks:

    No.16 stainless line leaked in two places on passenger side

    leak ex conveyor motor
    Electrical:
      

    Missing male to male adapter for the Ethernet cable.

    No.2 level sensor had a short at connection

    fans not working in control box/ bad fuse
    Mechanical:
     

    bolts came loose on radiatior mount
    paint touched up wrong shade

     engine had to be reflashed

    return idlers set wrong

    bolts came loose on radiatior mount

    landing leg front  dirve gear dogs not lined up
     
    Yes I have created Pareto charts of these defects (at least the categories-electrical was the most common), and we are going to accompany the inspector on some inspections to formulate a process rather than just a check-list. Remember, we have one assembly inspector and one in welding. Along with the lead person, that is our inspection department, along with a weld quality inspector. You are right, these aren’t random. I have many other of the same types of similar defects listed.
    Thanks again everyone-I am going to use these ideas to forge a new plan of action to attack this problem cycle. Although I wear the Lean hat here we get involved in quality as well, which is why I started this posting. Any other news please pass it along, I am very grateful, and a little humbled by your warm response!
    Chuck

    0
    #182653

    Vallee
    Participant

    While it is embedded in most of the posts, I did not see any human engineering questions or target assembler questions. Are there areas with defects in cramped areas… are there special tools being modified to get the job done… are defects increasing during peak times? Are parts having to be removed to install late parts? Are items out of sequence? Just because it may have been done this way for years does not make it efficient. While it is a custom build job their are basic skills and task that need to be observed to help.

    0
    #183325

    pedro
    Participant

    Hi Chuck !
    I was wondering how this problem ended up..
    If possible please let me  (us) know your results,, would be a good learning topic.
    appreciate it.
    Pedro

    0
Viewing 14 posts - 1 through 14 (of 14 total)

The forum ‘General’ is closed to new topics and replies.