Kickstarter Update - The logistics nightmare that is Chinese New Year

TL; DR:  We will be running an additional 3-4 weeks late at this point; final projected ship date to be updated by the end of Feb.

I blogged about the impacts of Chinese New Year last year briefly, especially to a new business owner.  This year, taxes were a bit easier, thanks to my awesome wife and household CFO helping with doing the taxes. 

As you may have read in the last update, we released our design to our injection molder about 4 weeks ago; that was the last major hurdle for us in engineering this and bringing it to production.

Unfortunately, we are not the only ones trying to beat "Spring Migration".  Spring Migration, or Chunyun, is the largest human migration and happens every year leading up to Chinese New Year and up to a month after Chinese New Year.  Chinese New Year is based on the lunar calendar and this year landed about a week and a half early.  No doubt, every major western engineering firm scrambled to try to beat it, and the resulting work load overloaded just about every good shop in China.  This, coupled with abnormally bad weather causing severe travel delays, meant that most factories are still NOT back to operating at their full capacity.  Even for our vendors that are not based in China, Chunyun is still a major logistics headache, as so many of the downstream suppliers (such as mold base suppliers, etc) are based in China.

(To give you an idea how much of an impact this is:  The number of passenger journeys during Chunyun exceed the population of China, hitting 2 BILLION PEOPLE in 2006.  Now imagine that this travel is done by train and by other modes of public transport that is suddenly handling 50-100 times their normal traffic loads.  Now throw in a Noreaster or a major storm event.  There are no winners in this logistics challenge).

We are seeing across-the-board delays as a result of this, unfortunately - not just for OpenBeam projects, but projects at our day job as well.  Our injection molder only now starting work on mold designs, even though the files were submitted a few weeks ago.  (In their defense - they did NOT accept our initial PO and confirm the delivery dates, and we have full faith that they will take care of us - they do AWESOME work).  Our Rev 2 Brainwave Pro (formerly known as the Brainwave II) boards were supposed to be delivered the week of Feb 1st.  We are now told that we may see the boards at the end of Feb.  (This is then followed by a week or two weeks of testing, before we can release the production run of the 450 boards used for Kickstarter fulfillment).

So, there's the long version of the story.  We will not be able to make a firm delivery promise for all the units until A) we receive a confirmed T0 date for our injection molded parts and B) receive our rev 2 test electronics board and have a chance to verify functionality of the entire board.  Currently based on our best guesses, the early birds should ship the week of April 15th, but this is an estimate subjected to change until confirmation of manufacturing schedule.

However, this does not mean that we've been idling the past 4 weeks.  This is what we have been up to:

A)  Moving ahead with machined aluminum vertexes

We are moving full speed ahead with our aluminum vertexes.  The manufacturer for these parts is based in South Elgin, Illinois, and is not affected by Chinese New Year at all.  We used 3D Printing to create the first fixtures to test the machine programming and design of the part; now we are making the aluminum fixtures required to machine the components on a production scale.

B)  Test printer fleet

We are really reaping the rewards of designing the Kossel and Kossel Pro to have fully interchangeable parts at the subassembly level (and having almost identical firmware).  With our plastic parts delayed we are turning to the Kossel Reprap branch to get printer test time in. 

We will be releasing the Kossel Reprap design in the next 7-14 days.  The idea is that by the time the Brainwave Pros arrive, we'll have a fleet of about 10 printers and we'll push those machines hard to see if there are any show stopping bugs. The lack of control boards to perform full testing is a disappointment, but the original Brainwaves are still perfectly capable boards when paired with matching motors.  We will make another announcement when we design the OpenBeam Kossel Reprap printer.

(Below:  Test output from the OpenBeam Kossel Reprap, using the Brainwave 1 with properly matched motors.  Printed from this thingiverse object, 100 micron layer height.  Print sample courtesy Metrix Create:Space

C)  Documentation efforts

We have purchased a Stratasys Mojo, a low end commercial FDM system, for printing injection molded Kossel Pro parts for prototyping.  The parts you see on the Kossel Pro below are all built on this machine.  This is not a cheap investment, but we are at the point in the company's growth where such an investment can be justified.  (Another way to look at this is that if I had purchased this machine when the Kickstarter was funded, we would have been able to close that painful prototyping cycle on the development of the Kossel Pro and we wouldn't be this late!  We wouldn't have gotten as good of a deal on the machine though, but the amount we paid in outside prototyping costs way more than the steep "used machine" discount we got)  We expect that this machine will pay dividends when it comes to the work on the next generation multi-material Kossel, as well as expansion to the OpenBeam system - both slated for when Kickstarter delivery is complete.

It may seem a bit off for an open source 3D printer developer to buy something as proprietary as a Stratasys Mojo, but the truth is, these machines are the only ones capable of repeatedly and reliably printing parts that require soluble support material.  I am paying through the nose for the reliability of being able to hit "print", go to work and come back to functional parts.

For now, the Mojo is churning out copies of the injection molded parts, so that our manufacturing engineer can start writing the assembly documentation.  By the time this project is completed we estimate that we would have spent 15 minutes in testing and 15 minutes in documentation for every hour of engineering time that went into the final product.  

D)  Kitting of subassembly components

In parallel with our documentation efforts, kitting is already under way.  We have purchased, and received majority of the 'off the shelf" components for fulfillment of the early bird kits.  (Some of these are risk buy items for evaluating supplier quality, when we qualified the part the parts then transfer to inventory for the early birds).  Fastener kitting had been the most painful of this (t's not a kit if it's missing a single screw, and it would be highly unfortunate if we had to rework kits because a handful of screws were missing).  There is much work to be done, and everything is subjected to careful checking.

E)  Test, test, test.

As we continue down the development path, testing is paramount.  We test our own documentation and processes -  we sit people down at an empty table and hand them fastener packs that have been kitted according to our work instructions to verify fastener count.  We use weight scales to check the aforementioned fastener pack and we use statistical process control to check the kitting of individual fastener packs.  We gutted a toaster for the resistive heater coils to stress test our power supplies, to make sure that they will hold up under the use of a heated bed (Dad didn't want to pay $80.00 for power resistors big enough to do the job safely, so he found a $5.00 toaster from Goodwill and hacked something up in the garage).  We use purpose build fixtures to control the cutting of all stock that have to be cut to length, from timing belts to PTFE insulators for the thermistor on the hot ends, to all the precrimped cable harnesses and bowden feed tubes.  

Oh, and one more thing...

When the OpenBeam Kossel Reprap is released in the upcoming 2 weeks, we are also releasing the tooling and calibration fixture designs as well, under open source license. We designed all the tooling and calibration fixtures (shown below) for building these printers to be printable on hobbyist Reprap class machines, to help spread the Delta love far and wide.  The tooling, no surprise, is also compatible with building Kossel Pros.

And, finally, the promised "More frequent updates"

We're been taking notes (see screenshot below).  Unfortunately, with all the supply chain fires, we have not had a lot of editorial time.  Priority one is still trying to not slip the delivery date too much.  Hopefully as the workload gets delegated off my shoulders, I'll have more time to write in the next few weeks.


-=- Terence and the Kossel Pro team