Print It Anyway: Want to Experiment More with 3D Printing? Tell us to PIA!

Discussion in 'General Discussion' started by 7943_deleted, Feb 26, 2014.

  1. 7943_deleted
    7943_deleted Member
    Brian123, great questions!

    To answer:

    1. Good question - the idea with PIA is that you accept the possibility of an imperfect model in exchange for extra information . When you ask about print layering issues or color issues - do you mean as determined by us ...or by you after you receive your model? It's a good question and one worth exploring more.

    2. About the price: Yes the price is the same, the risk you take is by forfeiting your refund. Also, this is an interesting question because traditionally we have assumed this price risk (in having to reprint a model 4 times before we got a successful one). With your idea about a discount for iteration...that's a great suggestion! Duly noted!

    3. Yes, we put your model through the entire production process, so even if it breaks we will continue cleaning and packing it as normal, we'll just also include feedback about where in the process it broke and why. Your order goes through the same steps that all regular orders go through and will be handled with the same care. In rare cases with polished or dyed materials, you may receive a product that hasn't gone through polishing or dyeing because it wouldn't have survived these processes. To enable you to learn and improve your design, you'll receive your product in the most complete state possible.

    Best,
    Natalia
     
    Last edited: Feb 27, 2014
  2. pete
    pete Shapeways Employee CEO
    Hi all,

    thanks for your feedback as always! And also thanks Paul, Tom, and others, for your patience hanging in there with us for all this time.

    We are really excited to offer the Print it Anyway feature, because it offers us a way for you to indicate what you want at scale.

    Yes it is true, 3 years ago it was easier to get feedback on your model, how to improve and make it printable. Over those years
    the amount of models we are assessing and printing has gone up by amazing amounts. We were looking at less than 10,000 models a month
    3 years ago vs over 100,000 today. This is 200 models a day vs over 2,000 a day. As a result the team has grown and we now have multiple people in 2 factories checking the models. We also offer post production services like polishing and dying, which add complexity to the manufacturing. To make the not-exact process of predicting whether we can reliably manufacture a product completely reproducible is HARD. Especially since we can't simply try a few times. Remember most products are all different. Also to keep the costs reasonable, which has always been our focus, it was also hard to have conversations around all parts we were "rejecting". As a result the experience suffered and to be clear we do not like that.

    Print it Anyway is another step we are taking after Wall Thickness Visualization and improvements in the checking process and others yet to come to improve the experience for you, and scale at the same time, keeping cost in mind.

    We definitely want to figure out how to get you all the models you want and communicate clearly when we can't and why. You have our commitment we are working on just that. Your feedback is helping us to get there. Thank you for that.

    Pete
     
  3. Brian123
    Brian123 Member
    Thank you for your answers, I will be interesting to see how this program and future ones develop.

    Expanding on Question 1

    I mostly print FCS, and often there are some quality issues that make it past inspection and shipped to me or a customer. So I'm worried these issues will still get through with PIA and void the usual reprint or credit. Maybe as part of the added communication when using PIA, clear photos can be taken and shared before shipping the model, or always flag for better inspection

    1: Badly calibrated layers
    A friend ordered 5 copies of this model at the same time, all packaged very differently (different sources?), and one was clearly below the quality of the others, showing badly aligned layers. Not a problem with the model, but was approved and shipped to the customer. Shapeways reprinted it.

    jagged-layers.jpg


    2: Non Uniform Color

    Below is a recent mild case from a customer who ordered this cat model. One side the color is more vivid than the other, making it too green and uneven. I had a similar issue in the past where the color slowly turned B&W on one side. My print was approved for a reprint, while my customer was told they had no control over this and was not issued a refund. This seems like an issue with printer maintenance, like a clogged line with the colors or print head. That being said, this issue was not as noticeable as mine that turned black and white. The customer did not receive a print at the same quality with uniform color that I featured in my store and examples shown below next to her print. Color can vary in richness, which is acceptable to an extent, but it should be uniform and not noticeably change across the model. So I was disappointed she did not get a refund.

    Again, this is a mild case compared to B&W, but the same non the less
    color-variation.jpg


    3: Support material removed and fused with model

    Here is one example where some support material wasn't removed before fusing with glue. Looking between the two front legs, there's lump of support material fused into the model. It was possible overlooked because it was white on white, but not a problem with the design and would of been bad if over a different color.

    On a recent print of the Lemur character, as seen in example 1 above. There was an obvious thin layer of white stuff fused on the tip of the black nose. Luckily it was easy to get to and thin enough to carefully rub and scratch away for a perfect print, but was nerve racking. With this example below, you can't do that

    support-material.jpg


    So those are all some kind of problem caused by a printer issue or cleaning issue, but still made it to me or the customer. If I used PIA with any of those, would I have been out of luck? These are just a few examples I had to show off hand, but you are making a risky print that survives the process, you hope it will not have secondary issues like these.
     
    Last edited: Feb 27, 2014
  4. Vidalcris
    Vidalcris Member
    Is this something cool ? maybe ... ;)
    For sure i will never use this option :p
     
  5. stop4stuff
    stop4stuff Well-Known Member
    Hi Pete,

    Thanks for your input.
    I'll go back to my OO scale phone boxes as an example.
    First off, they printed and were delivered just fine in FUD, no issues or reported problems.
    Then when the model 'got popular', too thin, can get damaged in cleaning, can get damaged during transit rejections started occuring.
    So yep, increasing business 10 fold can cause a rush and maybe standards of care for our models dropped which could explain the reasoning behind the rejections... however to me that is an undocumented change in business practice, i.e. 'wow, we got busy, and we've got less time, lets rush... Oh that got damaged because we rushed the cleanup or stuffed it in the wrong sized bag too vigorously', so the model subsequently gets marked for rejection in the future.

    This then leads me to the question, how will PIA change possible damage due to rough handling after printing?

    Personally I would like to get my hands on a ProJet 3500 & associated cleaning kit, to find out how the process works out if models are treated with care & respect.

    Cheers,
    Paul
    [hr][hr]
     
  6. MrNib
    MrNib Well-Known Member
    Also how about adding alumide to the list?
     
  7. 7943_deleted
    7943_deleted Member
  8. MitchellJetten
    MitchellJetten Shapeways Employee CS Team
    Get yourself a ticket to Eindhoven :)
    I'll be happy to show the printers!
     
  9. stop4stuff
    stop4stuff Well-Known Member
    three factors stopping me Mitchell;
    1 - I've no passport
    2 - #1 son is getting married in July = little money to spare
    3 - #1 daughter is 18 in July = no money to spare
    Plus I'd rather have a ProJet 3500 to play with in my own time = abosolutely no money and loads of fun :)

    Cheers,
    Paul
     
  10. 3rdboxcar
    3rdboxcar Member
    I had an vague idea that you try to print models more than once if a print fails, is there any way of us getting to know when this happens.

    I certainly would like to know if a model is on the borderline and it may take a few minor tweaks to make it rock solid.

    PIA is another step forward .... thanks
     
  11. HOLDEN8702
    HOLDEN8702 Well-Known Member
    First of all, I have to say PIA is a step forward in a good direction. And don't forget you haven't obligation to use it if you don't like it.

    Last week I asked to Shappy Team for something like this with a F.C.Sadstone model - They said yes, and I'm waiting for the results.

    All the improvements in the web are good. Wall thickness visualizer is the more practical and an awesome help to designers.

    But I can't understand why if a model pass the engineer check and goes to production, but passing a week the model is rejected as unprintable.

    Who is the weakest link of the chain: The fat fingers designer (me), the checking engineer (with his magnifying glass and big axe) or the manufacturing team?
     
  12. MitchellJetten
    MitchellJetten Shapeways Employee CS Team
    In this case the production team did print your model, but possibly broke and thus got rejected :(
     
  13. 7943_deleted
    7943_deleted Member
    HOLDEN8702, our goal with all of these projects is to get to place where there is no weakest link ;-)

    Between wall thickness checking, Print It Anyway, continuous calibration in our factories and a whole host of other projects in the pipeline, I'm pretty sure we'll get there soon!

     
  14. HOLDEN8702
    HOLDEN8702 Well-Known Member
    At the end, no results. Rejected again. My robots stood still in RoboHeaven.....
     
  15. JMorgen
    JMorgen Member
    Hey Paul

    I'm truly sorry for your prior issues with FUD quality. We've made sure to hire technicians that have a gentle demeanor with a passion for customer satisfaction. They know that the items they handle have emotional value for the person who has ordered them, and they try their best. Still, they are humans and they make mistakes at times. We have been reading the forums and working with Customer Service to identify ways to make our practices better. Currently, we are exploring and testing better packaging techniques to ensure your items make it to their destination intact.

    All teams that handle products associated with "Print It Anyway" have been instructed that they are to handle and care for all models equally. However, "Print It Anyway" models are actually receiving treatment beyond that of the non "Print It Anyway" models. If a model prints and makes it through shipping successfully, we are processing it as usual. If it does not, and the "Print It Anyway" feature is checked, we now have to perform an investigation as to why the quality failure occurred so that we can first identify whether the quality failure was a fault of our own or to give the best possible feedback to the customer.

    During this investigation, if we find that the quality issue occurred through fault of our own, we will correct this issue and try to produce your model with better quality before shipping. The only reason that we would intentionally ship to you with suboptimal quality is if the issue was unavoidable due to the design.

    Thanks for your feedback. I hope that we can improve our practices to support the needs of you and your shop.
     
    Last edited: Mar 1, 2014
  16. UniverseBecoming
    UniverseBecoming Well-Known Member
    Well I'll be darned! There may be hope for Shapeways after all! HAHAHA! :D

    Can we get revenge by ordering up our previously rejected models so we can prove to the rejection mongers that we were right all along?!! HAHA! :D

    I'm happy to see that Shapeways is making big moves to improve. :)
     
  17. PiecePrints
    PiecePrints Member
    any way ceramics will be on that list soon?
     
  18. 7943_deleted
    7943_deleted Member
    Blester, as this is a pilot program, we're keeping it to our three most popular materials, and the ones that we do the most of in-house. As we learn and see how PIA is going, we may be able to expand it to other materials.

     
  19. Innovo
    Innovo Member
    This would benefit me greatly since I prototype a lot . Problem is I use Stainless steel mostly so...

    BRING ME THE METAL!
     
  20. tebee
    tebee Well-Known Member
    Pete

    The point I was trying to make before, is that feedback from you to us, when you have problems producing our models, would help us to produce models which did not give you problems.

    While I applaud this initiative, I know you sometimes have to retry making things, if it still fails we get to hear about it, but if it succeeds on the second or subsequent try we don't.

    If it's something marginal in the model that causes this, it might be possible for us to redesign said model. These marginal models must add your costs and you must have existing systems in place to reprocess them. Would it be so difficult to add us designers in the the normal processing loop as well as is this case? Or is this something you plan after you have the experience from this initiative ?

    I've noticed that often models of mine go into production, but then return to a processing state - is this an indication you have had problems producing it ?

    Tom