WEBVTT

00:00:00.300 --> 00:00:04.500
Welcome to Wintergatan Wednesdays
where I'm trying to finish the Marble Machine X.

00:00:04.900 --> 00:00:08.200
And the plywood parts lying on my work table today

00:00:08.201 --> 00:00:13.201
are going to play a crucial role
in me being able to finish this machine or not.

00:00:13.400 --> 00:00:17.600
I love these pieces. They are really
tomato sauce umami pieces... [slurping sound]

00:00:17.601 --> 00:00:22.999
By the way, thank you so much for the amazing
response on last week's update video.

00:00:23.000 --> 00:00:25.200
That was amazing to see. Thank you.

00:00:25.201 --> 00:00:29.299
In today's video, we are going to
assemble these pieces,

00:00:29.300 --> 00:00:32.600
and try them out to see if they
work and do what they should do.

00:00:32.700 --> 00:00:36.800
We're also going to meet the amazing
Marble Machine X CAD team.

00:00:36.801 --> 00:00:39.299
But before that, let's get to some assembly.

00:00:39.300 --> 00:00:45.600
When designing these pieces, I've been using one of my
favorite trick: I've split the design in the middle.

00:00:45.800 --> 00:00:51.100
So if you check here, if I take this
piece and I turn it over 180 degrees,

00:00:51.101 --> 00:00:53.201
and I put it together with the other piece.

00:00:53.202 --> 00:00:58.002
That creates a shape that is impossible
for me to cut on the CNC machine.

00:00:58.100 --> 00:01:01.700
And then I use the same technique
on these crooked marble paths.

00:01:01.800 --> 00:01:07.200
So if I take these two pieces, they are mirrored,
and I put them together, like so.

00:01:07.700 --> 00:01:11.900
We have a crooked marble path
inside this plywood block.

00:01:12.200 --> 00:01:17.500
This marble module will alter
the length of the marble paths,

00:01:17.700 --> 00:01:22.300
controlling that the marbles are flush
on the top of the marble divider.

00:01:23.400 --> 00:01:30.900
I've designed these so you can open and close
to keep the different modules locked in place here.

00:01:30.901 --> 00:01:33.301
[drill whirring]

00:01:42.300 --> 00:01:44.400
[ratchet clicking]

00:01:44.401 --> 00:01:48.991
So these are all unique marble
paths with different lengths.

00:01:49.100 --> 00:01:52.800
As you can see, I can just fit four marbles on there.

00:01:53.600 --> 00:01:59.700
This "6" here, you can see that it's crooked.
I've shifted the path, six millimeters...

00:02:00.000 --> 00:02:05.400
You see that space there? That's the whole reason
for the existence of this part. [chuckles]

00:02:05.700 --> 00:02:09.400
Let's check "14", even more space.

00:02:09.700 --> 00:02:12.800
And the longest path I have is this one, "22",

00:02:13.300 --> 00:02:17.000
and here you can see that I can even fit five marbles.

00:02:17.001 --> 00:02:22.401
So now when I know I have... the longest one can
accommodate a whole extra marble,

00:02:22.600 --> 00:02:25.700
then I know I have all the combinations I will ever need.

00:02:25.900 --> 00:02:30.799
So every path has its corresponding mirrored path...

00:02:30.800 --> 00:02:32.100
like so.

00:02:32.101 --> 00:02:37.201
These modules will later slot in here,
but first we have to assemble them.

00:02:37.400 --> 00:02:42.300
To align them perfectly, I'm going to use some clamps.

00:02:44.200 --> 00:02:48.699
[drilling sound]

00:02:53.600 --> 00:02:59.100
I hardly feel the edge with my finger here.
That's perfect!

00:02:59.500 --> 00:03:04.300
Looking at the outside of this module,
we don't know what curves are inside.

00:03:04.301 --> 00:03:10.301
So I'm gonna mark them with their number
on a surface that will be hidden when in use.

00:03:10.400 --> 00:03:14.700
I assembled all of them and they're gorgeous.
[chuckles]

00:03:15.300 --> 00:03:20.200
So, if there's something I've learned
during the Marble Machine X project,

00:03:20.201 --> 00:03:22.561
it's to machine plywood.
[laughs]

00:03:23.200 --> 00:03:24.880
Let's try them!

00:03:26.800 --> 00:03:32.300
So here we can clearly see how the top marble
sits lower and lower and lower.

00:03:32.500 --> 00:03:35.900
Let's try them on the docking station.

00:03:36.000 --> 00:03:37.500
This is how it works.

00:03:37.501 --> 00:03:44.901
It slides in like so,
and then I can take another one, put it like so.

00:03:45.700 --> 00:03:50.900
And I can change them out to adjust
the top height of the top marble.

00:03:51.500 --> 00:03:56.800
But, there is a radius from the
tool of the CNC machine, <i>here</i>.

00:03:56.900 --> 00:04:02.100
And here's a sharp corner and I want this radius here
to make these a little bit more sturdy.

00:04:02.300 --> 00:04:05.200
So now I need to take off this sharp corner.

00:04:05.201 --> 00:04:08.701
The radius of this fillet is three millimeter.

00:04:09.000 --> 00:04:13.500
So I'm scoring a three millimeter
line here with my caliper.

00:04:14.000 --> 00:04:19.600
And now I can use the disc sander to sand away
this corner until I come up to the scored line.

00:04:21.800 --> 00:04:25.000
Here you can see I rounded over the corners

00:04:25.001 --> 00:04:28.701
and now... it sits flush in the bottom.

00:04:30.000 --> 00:04:31.100
Perfect!

00:04:31.500 --> 00:04:36.300
All the pieces are ready and
I can lower these cam levers,

00:04:36.800 --> 00:04:40.600
so we can test this out and see how it fits.

00:04:40.601 --> 00:04:45.201
This thing is actually gonna go up here on the machine.

00:04:45.202 --> 00:04:49.602
And I can already see a misalignment. Let me show you.

00:04:49.603 --> 00:04:54.203
I slide in the modules;
You can then see the misalignment.

00:04:54.204 --> 00:04:58.900
This should be concentric;
The parts are too much this way.

00:04:58.901 --> 00:05:04.301
So you can see the exact same effect
on the top row but even worse.

00:05:04.302 --> 00:05:07.702
Here you hardly can spot the edge of the top edge.

00:05:07.800 --> 00:05:14.000
I'm gonna put shim washers
both here and between them, here.

00:05:14.900 --> 00:05:20.600
3.14, so I need the two washers to be exactly pi.

00:05:20.601 --> 00:05:24.101
3.14, pi again.

00:05:24.102 --> 00:05:26.302
So you can use pi in this case,

00:05:26.303 --> 00:05:31.603
because whenever you take something
completely round, and you put it twice,

00:05:31.604 --> 00:05:35.204
the width will always be 3.14...

00:05:36.100 --> 00:05:37.500
It's very useful.

00:05:40.000 --> 00:05:41.000
[drilling]

00:05:41.300 --> 00:05:42.500
I made a maracas!

00:05:42.501 --> 00:05:45.501
[rhythmic rattle]

00:05:47.900 --> 00:05:49.800
[thumping]

00:05:50.400 --> 00:05:52.600
[drilling and rattling]

00:05:53.100 --> 00:05:55.100
That looks super centered,

00:05:55.400 --> 00:05:57.500
and this looks perfect as well.

00:05:57.501 --> 00:06:00.501
Washer stacks has saved us once again,

00:06:00.502 --> 00:06:03.102
and we can now test this thing for real.

00:06:03.300 --> 00:06:06.400
These two PMMA pipes are just a visualization.

00:06:06.600 --> 00:06:10.300
Think of the top of these pipes
as the top of the marble divider,

00:06:10.301 --> 00:06:13.001
where the top marble needs to be absolutely flush.

00:06:13.002 --> 00:06:18.702
So we're gonna start with "0" and "2",
just to have something to experiment with.

00:06:18.703 --> 00:06:22.403
So I slot this in like a server,
this I think is so cool,

00:06:22.404 --> 00:06:28.204
like: shtuck, shtuck, shtuck,
and then we're gonna fill the pipes.

00:06:29.800 --> 00:06:35.700
Do you see that this marble is sticking up,
and that this marble is sticking up, but much less?

00:06:35.701 --> 00:06:39.401
With my modular system, I can now tweak this.

00:06:39.402 --> 00:06:43.002
So let's start over here with this channel.

00:06:43.400 --> 00:06:47.300
It needs to go down... quite a lot.

00:06:47.301 --> 00:06:49.501
So I'm going to take this one out...

00:06:49.700 --> 00:06:51.900
without dropping all the marbles.

00:06:52.000 --> 00:06:56.000
So we had "2" and now I'm inserting "16".

00:06:59.300 --> 00:07:01.000
Closing...

00:07:02.200 --> 00:07:04.000
Look at that!

00:07:04.001 --> 00:07:06.701
Almost exactly flush!

00:07:06.702 --> 00:07:10.500
I actually want this top
marble to be a little bit higher.

00:07:10.501 --> 00:07:14.301
We used "16", let's try "14".

00:07:14.900 --> 00:07:16.600
This is a one-time setup.

00:07:16.601 --> 00:07:20.301
I would never have to do this
more than once on the real machine.

00:07:21.700 --> 00:07:24.699
Yeah! Boom, that's perfect!

00:07:24.700 --> 00:07:29.286
This one is just a little bit over. Number "6", here.

00:07:30.400 --> 00:07:35.400
[marbles falling and clattering
on the table and on the floor]

00:07:42.100 --> 00:07:44.100
Slot this one in...

00:07:44.700 --> 00:07:46.499
Boom, it's flush!

00:07:46.500 --> 00:07:47.500
[poosh]

00:07:47.501 --> 00:07:52.899
This part takes care of a function
to make the marbles flush all by itself,

00:07:52.900 --> 00:07:57.900
leaving the new marble gates
free to be much simpler.

00:07:57.901 --> 00:08:02.101
I make all the subassemblies
more stupid, more reliable.

00:08:02.102 --> 00:08:05.902
If I keep on doing that,
the Marble Machine X will actually work.

00:08:06.500 --> 00:08:11.099
What is more amazing though, is
what's going to happen now, because...

00:08:11.100 --> 00:08:13.999
while I am building in my workshop,

00:08:14.000 --> 00:08:18.899
we have a CAD team working on the
CAD model of the Marble Machine X.

00:08:18.900 --> 00:08:21.700
It's been a huge disappointment for me personally,

00:08:21.701 --> 00:08:27.301
that I haven't been able to keep...
update the CAD model of this project.

00:08:27.400 --> 00:08:34.400
So while back I asked people on the discord server
if they wanted to be part of a CAD team.

00:08:34.401 --> 00:08:37.901
This project has been going on in the background
for the last four months,

00:08:38.000 --> 00:08:41.799
and it's just filled with amazing, amazing people.

00:08:41.800 --> 00:08:46.599
So right off the bat, I just want to say thanks to
everyone who are part of the CAD team.

00:08:46.600 --> 00:08:52.200 line:1
And especially, you're gonna meet Maureen
from Meticulae Designs youtube channel.

00:08:52.400 --> 00:08:55.500 line:1
Maureen is the defender of the Kinetic Fingers,

00:08:55.501 --> 00:09:01.301 line:1
and I asked Maureen if she could make
a little video introducing the CAD team.

00:09:01.400 --> 00:09:05.600 line:1
So after this successful try,
I'm gonna hand it over to Maureen.

00:09:05.700 --> 00:09:08.500 line:1
Check out her youtube channel, Meticulae Designs,

00:09:08.860 --> 00:09:10.420 line:1
and... enjoy!

00:09:11.040 --> 00:09:12.199
[large gong sound]

00:09:12.200 --> 00:09:16.700
Maureen: - Three months ago...
a butterfly made a list.

00:09:18.800 --> 00:09:20.200
[large gong sound]

00:09:20.201 --> 00:09:25.101
Three months ago...
the butterfly put out a call for help.

00:09:26.500 --> 00:09:27.500
[large gong sound]

00:09:27.501 --> 00:09:34.001
And three months ago... the call was
answered by an army of volunteers.

00:09:34.500 --> 00:09:37.500
But did they know what they were getting into?

00:09:37.501 --> 00:09:39.059
[♪ unreleased/The Ondophone Prototype 1.1 ♪]

00:09:39.060 --> 00:09:44.399
Inside that innocent-looking simple little
row, a great beast was hidden.

00:09:44.400 --> 00:09:49.700
The beast was born from software
limitations, collaborator restrictions,

00:09:49.701 --> 00:09:52.701
and the abandonment of its tracking system.

00:09:53.000 --> 00:09:59.400
It threatened madness, despair, and
hopelessness on anyone attempting to tame it.

00:09:59.900 --> 00:10:03.300
But the thousand arms of the volunteers
were not deterred.

00:10:03.700 --> 00:10:08.700
They bared their spoons, because they knew
the secret to conquering this beast.

00:10:09.200 --> 00:10:11.500
How do you eat an elephant?

00:10:16.000 --> 00:10:22.000
[♪ unreleased/The Ondophone Prototype 1.1 ♪]

00:10:27.600 --> 00:10:31.500 line:1
Hi, I'm Maureen! I'm here to talk
about the CAD MMX Project

00:10:31.501 --> 00:10:34.701
and share what the CAD team's
been up to in our first months.

00:10:35.200 --> 00:10:40.700 line:1
When Martin announced this project on discord,
dozens of CAD modelers immediately jumped on board.

00:10:41.500 --> 00:10:45.999
By now we have over 200 people signed
up to follow the CAD team discord channel.

00:10:46.000 --> 00:10:48.600
But finding volunteers turned out
to be the easy part.

00:10:49.500 --> 00:10:56.200
The difficulty in this project lies with the complexity
of the CAD model, pushing Fusion 360 past its limits.

00:10:56.201 --> 00:11:02.399
And with the management of thousands of parts
designed by multiple people over a number of years,

00:11:02.700 --> 00:11:07.500
many of which have been physically changed
on the machine, but not updated in CAD.

00:11:07.800 --> 00:11:11.000
And the only reliable resource
we have to track all of this

00:11:11.000 --> 00:11:13.500
are the MMX build videos Martin's made.

00:11:13.500 --> 00:11:17.549
So Martin wasn't kidding when he said that
the biggest challenge to this project was

00:11:17.550 --> 00:11:20.799
finding a project manager,
because instead of just one,

00:11:20.800 --> 00:11:24.210
we currently have six managers of the CAD team.

00:11:24.400 --> 00:11:27.400
Some of them have been members
of the e-team for quite a while,

00:11:27.401 --> 00:11:29.401
while others are new to the team.

00:11:30.200 --> 00:11:33.600
Tim Keller's our project leader
and liaison with Martin.

00:11:33.601 --> 00:11:38.101
Tim Mitchell, Will Francis, and David Lewis
are our Quality Control team.

00:11:38.102 --> 00:11:41.399
Tim &amp; Will make sure the parts
our team members are working on

00:11:41.400 --> 00:11:45.200
match the modeling and defeaturing
guidelines set forth in our instructions,

00:11:45.400 --> 00:11:50.100
and David oversees assembly of the approved
parts into our new Master Assembly.

00:11:50.400 --> 00:11:52.400
Jordan Bowden is new to the e-team.

00:11:52.401 --> 00:11:55.401
He's an Airtable wizard who built our new database

00:11:55.402 --> 00:11:57.499
and manages all our assets.

00:11:57.500 --> 00:11:58.900
And I'm Maureen Krauth.

00:11:58.901 --> 00:12:03.801
I'm new to the e-team too, and my job is maintaining
our documentation and People Management.

00:12:05.899 --> 00:12:09.799
The project is broken into two phases.
[♪ unreleased/Testing E-Bow on Ondophone ♪]

00:12:09.800 --> 00:12:13.949
In Phase 1, we're simplifying
the CAD model inside Fusion 360

00:12:13.950 --> 00:12:19.750
by removing all fasteners, defeaturing
internal geometry, and remodeling parts efficiently,

00:12:19.751 --> 00:12:23.599
to keep the 3D model
computationally light and easy to load.

00:12:23.900 --> 00:12:26.800
At the same time, we're bringing
the CAD model up-to-date

00:12:26.801 --> 00:12:29.401
so that it matches the physical machine.

00:12:29.402 --> 00:12:34.902
When Phase 1 is complete we will release the
CAD model to the MMX fan community for free,

00:12:34.903 --> 00:12:39.299
and if the ideas people are sharing on
discord already are any indication,

00:12:39.300 --> 00:12:42.800
there'll be a lot of cool things
coming out after this project is done.

00:12:43.900 --> 00:12:49.000
In Phase 2 we will be rebuilding the CAD MMX
from scratch, likely using another program.

00:12:49.100 --> 00:12:53.099
We will closely scrutinize each and every
part, and redesign everything we can

00:12:53.100 --> 00:12:57.700
to be more easily manufacturable, so Martin can
have backup parts for the World Tour,

00:12:57.701 --> 00:13:01.901
that don't require improvisations or manual CNC-ing.

00:13:03.000 --> 00:13:07.500
Along with these phases, we're recreating
the PBS inside a program called Airtable,

00:13:07.501 --> 00:13:09.301
to track everything we're doing.

00:13:09.302 --> 00:13:11.602
This new database has the old one in it,

00:13:11.603 --> 00:13:14.049
along with other data crucial for our team,

00:13:14.050 --> 00:13:17.949
including time-coded reference for which parts
are seen in which episodes,

00:13:17.950 --> 00:13:21.299
and flowcharts mapping out the
power train and marble flow.

00:13:22.400 --> 00:13:28.500
Ultimately we'll create a bill of materials that tracks
every single part and piece of hardware for Phase 2,

00:13:28.501 --> 00:13:31.001
with details for machining and assembly.

00:13:32.720 --> 00:13:34.949
[♪ unreleased/Cyber Bass Funkalizer ♪]

00:13:34.950 --> 00:13:39.150
To start this project, Martin and Tim Keller
went through the whole machine together,

00:13:39.151 --> 00:13:43.199
sorting the machine into roughly
100 logical assembly groupings,

00:13:43.200 --> 00:13:46.099
which make more sense now
that the machine's almost done,

00:13:46.100 --> 00:13:50.400
than the older organization of the
design-as-you-go PBS system.

00:13:50.800 --> 00:13:56.300
Martin's done walk-around videos for us, explaining
what's included in each of the new assembly groupings.

00:13:56.800 --> 00:14:01.599
We then dig through the old e-team folders
in Fusion to find any existing CAD models,

00:14:01.600 --> 00:14:04.599
and bring them into our new team project.

00:14:04.600 --> 00:14:09.300
Any part that's been changed from its
original design will need to be updated,

00:14:09.301 --> 00:14:13.701
while other parts were never designed in CAD
and have to be modeled from scratch.

00:14:14.400 --> 00:14:17.900
Every week on Wednesdays we
have a team meeting inside discord

00:14:17.901 --> 00:14:23.300
where we go over the parts we'll be assigning, and answer
questions people have about what they're working on.

00:14:23.500 --> 00:14:28.200
And then we watch the Wintergatan Wednesday
video together as a team. Hi guys!

00:14:28.201 --> 00:14:33.501
And collectively freak out any time Martin gets out
the angle grinder for a part we've already done.

00:14:34.100 --> 00:14:36.900
Martin: - Time for the angle grinder.
CAD team members: - Ooh... [chuckles]

00:14:37.400 --> 00:14:40.800
Tim Keller: Another part goes from
completed back to not completed.

00:14:40.801 --> 00:14:42.801
Andrey Shur: - Yeees!
[laughs]

00:14:44.500 --> 00:14:47.700
Andrey Shur: - Angle grind me, Martin!
[laughs]

00:14:50.200 --> 00:14:55.200
Our volunteers span the entire globe,
speak 19 different languages,

00:14:55.201 --> 00:14:59.801
and have a range of experience levels
in Fusion 360 and other CAD programs

00:14:59.802 --> 00:15:03.022
from absolute newbie to experienced master.

00:15:03.200 --> 00:15:08.399
Because there are only so many parts to go around
and we want to try to include everyone,

00:15:08.400 --> 00:15:14.600
we've come up with a partnership system where we assign
more experienced Seniors with less experienced Juniors,

00:15:14.601 --> 00:15:19.201
and team people up based on how their assemblies
connect to each other inside the machine.

00:15:19.300 --> 00:15:21.700
These groups are overseen by team leaders,

00:15:21.701 --> 00:15:24.988
who look for alignment issues
as they test-fit the parts together,

00:15:25.000 --> 00:15:30.199
and we also have a Status Trackers team that keeps
the organizational side of things running smoothly.

00:15:30.500 --> 00:15:35.899
They ensure everything is kept updated
between discord, Fusion, and Airtable,

00:15:35.900 --> 00:15:39.800
so we can track the status
of every part of the model at any time.

00:15:40.400 --> 00:15:46.300
So far, 74 team members have been assigned
to work on 80 different parts of the CAD MMX.

00:15:47.300 --> 00:15:51.300
Once the assignments have been announced,
the volunteers figure out if those models

00:15:51.301 --> 00:15:54.401
match the video reference or need to be updated.

00:15:54.402 --> 00:15:56.801
Martin measures specific bits of the machine

00:15:56.802 --> 00:15:59.702
that our team members request for accuracy.

00:15:59.703 --> 00:16:02.299
They update their models, defeature them,

00:16:02.300 --> 00:16:05.700
and submit them to be checked by Quality Control.

00:16:05.701 --> 00:16:08.898
After enough parts in an assembly group
have been approved,

00:16:08.899 --> 00:16:11.799
they're ready for the team leaders
to begin Assembly Check.

00:16:12.300 --> 00:16:16.600
Aligning the power train to the base frame
was the first step in this process.

00:16:17.400 --> 00:16:20.800
David's created a master skeleton
to check everything against,

00:16:20.801 --> 00:16:25.701
and the team leaders look for alignment issues
and figure out if everything's working together.

00:16:26.100 --> 00:16:28.900
The Assembly Check process can take a while,

00:16:28.901 --> 00:16:32.651
because everything is connected to
everything else, and each part needs to be

00:16:32.652 --> 00:16:36.352
checked against not only what it's mounted
to but also what it's moving with.

00:16:37.900 --> 00:16:40.799
Bite by bite, we're making good progress,

00:16:40.800 --> 00:16:44.100
and we'll keep working our way
all the way up to the top.

00:16:46.700 --> 00:16:49.900
Three months ago, none of this existed.

00:16:51.800 --> 00:16:57.600
Three months ago, Martin had no idea if people
would even want to tackle this monster project.

00:16:57.601 --> 00:17:02.601
And three months ago, that elephant
had no idea what was coming for it.

00:17:02.602 --> 00:17:08.602
[♪ unreleased/ Cyber Bass Funkalizer ♪]

00:17:18.350 --> 00:17:19.450
[chuckles]

00:17:19.451 --> 00:17:21.500
Martin: I'm laughing so much right now;

00:17:21.501 --> 00:17:24.001
I'm picturing the CAD team

00:17:24.002 --> 00:17:27.502
in these live-viewing parties of the episodes,

00:17:27.900 --> 00:17:30.900
like when I pull out the angle grinder
[laughs]

00:17:30.901 --> 00:17:32.501
and someone has, like,

00:17:32.502 --> 00:17:35.202
CADed the piece down to perfection

00:17:35.203 --> 00:17:37.103
and they just laugh about it!

00:17:37.104 --> 00:17:39.404
That says everything; That we're a...

00:17:39.405 --> 00:17:41.549
huge club of delayed gratification.

00:17:41.550 --> 00:17:45.399 line:1
So, check out Maureen's videos on the Kinetic Fingers,

00:17:45.400 --> 00:17:47.600 line:1
and thank you, Maureen, for making this video.

00:17:47.601 --> 00:17:52.101 line:1
My favorite part of the CAD project is the Phase 2,

00:17:52.102 --> 00:17:56.402 line:1
where we're actually going to redesign
all the mistakes on this machine. [chuckles]

00:17:56.403 --> 00:18:00.103
Where we can, like, together,
me and the whole CAD team,

00:18:00.104 --> 00:18:04.004
everyone can, like, pitch in
ideas for simplifications.

00:18:04.005 --> 00:18:08.004
I'm thinking, for example,
I wanna skip the magnet lifts. [chuckles]

00:18:08.005 --> 00:18:12.599
They're causing so much trouble. Everyone told me
from the beginning: "don't do magnet lifting".

00:18:12.600 --> 00:18:14.799
I wanted this cool visual effect.

00:18:14.800 --> 00:18:21.099
I think we should redesign the MMY to have
a classical marble machine lift technique.

00:18:21.100 --> 00:18:23.500
We can then skip the whole demag wheel.

00:18:23.501 --> 00:18:26.999
Sorry, you who have CADed the demag wheel!

00:18:27.000 --> 00:18:32.999
But basically, the MMY is going to be, like,
the machine as I should have done it right.

00:18:33.000 --> 00:18:36.499
So, the original Marble Machine
became a prototype for this.

00:18:36.500 --> 00:18:38.999
This is the prototype for the MMY.

00:18:39.000 --> 00:18:42.100
So, potentially, on the world tour,

00:18:42.101 --> 00:18:45.899
I will play live on the MMY. That can happen.

00:18:45.900 --> 00:18:51.799
At the main rudder of this gigantic
cruise ship is Tim Keller.

00:18:51.800 --> 00:18:55.200
So, you know him from
Tim Keller and The Broken Endmills.

00:18:55.201 --> 00:18:59.101
Apparently, they changed name of the
band to The Elephant Eaters. [laughs]

00:18:59.102 --> 00:19:03.102
Tim Keller has been a huge part
of the whole Marble Machine X project,

00:19:03.103 --> 00:19:08.903
and is now heading this CAD team
and is doing an awesome job out of it.

00:19:08.904 --> 00:19:13.603
Let's check in on the latest update on the
CAD project from Tim Keller himself.

00:19:13.604 --> 00:19:18.399 line:1
Tim: The update on the CAD project is
"my team rocks". My team is... is so cool.

00:19:18.400 --> 00:19:23.100 line:1
It makes it easy to be the leader
of a team when your team just...

00:19:23.101 --> 00:19:26.501 line:1
is all self-starters,
and we're working on something where...

00:19:26.502 --> 00:19:28.702 line:1
everyone, like, really wants to be part of it.

00:19:28.703 --> 00:19:31.903 line:1
It makes leading a team like that
really easy and really fun.

00:19:31.904 --> 00:19:34.804 line:1
There's so many people who are
part of the discord community.

00:19:34.805 --> 00:19:38.259 line:1
I've been able to go and find people who
look like... "Oh, this person is pretty engaged",

00:19:38.260 --> 00:19:41.699 line:1
"this person has a knowledge base or a skill base
that I think might be interesting".

00:19:41.700 --> 00:19:43.700
Have a couple of conversations with them,

00:19:43.701 --> 00:19:47.901
bring them onto the team, and all of a sudden they're
doing amazing things that I wouldn't have thought of.

00:19:47.902 --> 00:19:48.602
And I love it.

00:19:48.603 --> 00:19:53.799
Early in the Wintergatan series, you guys
had an episode on the PBS system, right?

00:19:53.800 --> 00:19:55.900
That Chris had developed with Martin.

00:19:55.901 --> 00:20:00.001
This is the PBS system, and the PBS
system was basically a number line, right?

00:20:00.002 --> 00:20:03.002
So as the numbers went up,
you were able to categorize

00:20:03.003 --> 00:20:07.199
and get all the different parts
kind of into groups along that number line.

00:20:07.200 --> 00:20:10.599
This is a database, and the company Airtable

00:20:10.600 --> 00:20:12.949
actually gave us this for free,
which was awesome of them.

00:20:12.950 --> 00:20:15.699
They didn't ask for anything in return,
which was super awesome.

00:20:15.700 --> 00:20:18.949
And you can go down all the way
as deep as you want on these things,

00:20:18.950 --> 00:20:22.450
all the way back to, let's see...
vibraphone resonator pipes.

00:20:22.451 --> 00:20:26.299
You can see, these are episodes...

00:20:26.300 --> 00:20:30.000
where he's talked about the vibraph...
the resonator pipes.

00:20:30.001 --> 00:20:34.901
Here are... time stamps, transcriptions...
[laughs]

00:20:34.902 --> 00:20:38.602
of what's happening in that episode.
I mean, it...

00:20:38.603 --> 00:20:41.249
And this isn't... this isn't all just the software,

00:20:41.250 --> 00:20:44.450
we have a team of people who are doing this,
and putting this information in, right?

00:20:44.451 --> 00:20:46.151
Garbage in, garbage out.

00:20:46.152 --> 00:20:50.099
The fact that this community is doing this, and...

00:20:50.100 --> 00:20:53.300
and especially, you know, as
someone who gets to lead this team,

00:20:53.300 --> 00:20:56.800
I actually don't know who's doing this, right?

00:20:56.801 --> 00:21:00.099
This is being managed by someone
who Maureen is managing,

00:21:00.100 --> 00:21:05.700
And it's all taking care of itself, because...
because the community is so just energetic

00:21:05.701 --> 00:21:08.401
and willing to be... to play some small role in it.

00:21:08.600 --> 00:21:11.499
It's... super cool.

00:21:11.500 --> 00:21:14.300
When we get into phase two of this project,

00:21:14.301 --> 00:21:19.699
we're actually going to be doing it using...
a master skeleton

00:21:19.700 --> 00:21:24.600
of the machine where we have
all the key components laid out.

00:21:24.601 --> 00:21:29.201
Where each axis of each, you know,
linear shaft has to be,

00:21:29.202 --> 00:21:34.502
where the mating points of
all the gears are gonna be,

00:21:34.503 --> 00:21:38.799
and where all of the different paths
that the marbles could take through.

00:21:38.800 --> 00:21:44.300
And we're gonna start with a skeleton like that,
and then design the machine around it.

00:21:45.800 --> 00:21:51.999
That's a big change to the way that Martin
originally designed it with Fusion 360.

00:21:52.000 --> 00:21:56.900
Fusion 360 is what's called a top-down modeling
environment, where you design individual parts,

00:21:56.901 --> 00:21:59.101
and then you can kind of break them up into assemblies.

00:21:59.102 --> 00:22:02.999
Things like Solidworks and some other ones
are bottom-up, where you can say:

00:22:03.000 --> 00:22:08.000
"Hey, this is the overall architecture of the...
of what you're building,"

00:22:08.001 --> 00:22:10.999
"and now let me make parts
that fit into that architecture."

00:22:11.000 --> 00:22:16.000
When you are making something as complicated as the MMX,
you really do need a bottom up environment,

00:22:16.001 --> 00:22:21.801
which is why in phase two, we're gonna be going to
a more professional software platform like that.

00:22:22.700 --> 00:22:27.400
So there are several kind of cool VR projects
that are out there. There are...

00:22:27.401 --> 00:22:32.401
One is a guy who actually did
his own simplified Marble Machine CAD.

00:22:32.402 --> 00:22:36.502 line:1
Brian Harms: Hi Martin, I have created a simplified
version of the MMX that you can actually play in VR.

00:22:36.503 --> 00:22:40.999 line:1
Ah! Forgot the crank downstairs,
let me go grab that. Oh no, Wilson...

00:22:41.000 --> 00:22:44.999 line:1
Sometimes when he drinks he tries to
CNC an army of evil Wilsons.

00:22:45.000 --> 00:22:47.800 line:1
All right, let's slap this on the machine.
[plop]

00:22:49.599 --> 00:22:53.799 line:1
[music slightly off rhythm]
[♪ Wintergatan: Marble Machine ♪]

00:22:53.800 --> 00:22:55.700 line:1
There it is. I hope you liked it!

00:22:55.701 --> 00:22:57.301 line:1
[strange laughter]
What was that?

00:22:57.302 --> 00:23:00.002 line:1
[high pitched Wilson noises]
Evil Wilson #1: Attack! Let's go!

00:23:00.003 --> 00:23:03.503 line:1
Brian: - Oh ho! Not again.
[Wilson laughs]

00:23:03.504 --> 00:23:06.704 line:1
Evil Wilson #1: Now you'll never finish this project!
Evil Wilson #2: Oh, sick burn evil Wilson!

00:23:06.705 --> 00:23:10.499
Tim: The model that we spit out
at the end of this will be de-featured,

00:23:10.500 --> 00:23:14.299
it'll have all of the relevant parts,
it will be cosmetically accurate.

00:23:14.300 --> 00:23:16.699 line:1
It'll be, you know, the real machine.

00:23:16.700 --> 00:23:20.500 line:1
And people will be able to do with that digital
asset, whatever the heck they want.

00:23:20.501 --> 00:23:22.901 line:1
There is so much you could do...

00:23:22.902 --> 00:23:27.802 line:1
We could make literally a playable
virtual reality Marble Machine.

00:23:28.500 --> 00:23:34.700 line:1
Yeah, Martin is always very thankful and very...
very keen to express his gratitude to these people,

00:23:34.701 --> 00:23:37.831 line:1
but we're all just doing it 'cause it's fun. And...

00:23:39.000 --> 00:23:45.600 line:1
I love that I get a place to play with
people who are so generous,

00:23:45.601 --> 00:23:50.001 line:1
and who are so smart, and just gifted
that the community itself is rewarding.

00:23:50.500 --> 00:23:56.400 line:1
I find myself just chatting with, you know,
Maureen and with Jordan and David Lewis, it's...

00:23:56.401 --> 00:24:01.999 line:1
We're our own family and if Martin got
hit by a Marble Machine, you know,

00:24:02.000 --> 00:24:05.999
two weeks from today, we'd probably
still be doing this, right?

00:24:06.000 --> 00:24:09.999
The community has its own rewards,
and I love being part of it.

00:24:10.000 --> 00:24:11.500
Martin: What can I say?

00:24:11.501 --> 00:24:13.399
[Martin singing]
[♪ Tegan and Sara: Everything Is Awesome ♪]

00:24:13.400 --> 00:24:15.999
♪ Everything is awesome, ♪
♪ everything is cool when you're part of a team. ♪

00:24:16.000 --> 00:24:18.800
My main intention with the discord server

00:24:18.801 --> 00:24:23.999
was for like-minded people to have a
platform where we could meet each other.

00:24:24.000 --> 00:24:26.299
That's exactly what is happening right now.

00:24:26.300 --> 00:24:31.199
And I mean, if there's any tech businesses
out there in the world looking for amazing talent,

00:24:31.200 --> 00:24:36.500
lurking around in the Wintergatan discord server
might be one of your best bets. [chuckles]

00:24:36.501 --> 00:24:39.700
I loved how you described the team members, Tim,

00:24:39.701 --> 00:24:43.299
and how many people are just like,
taking their own initiatives.

00:24:43.300 --> 00:24:48.199
The attacking Wilson's... [chuckles]
Brian, that was absolutely genius!

00:24:48.200 --> 00:24:51.699
I was wondering why my stash
of Dragon tears was empty...

00:24:51.700 --> 00:24:59.300
I mean, the sky's the limit when it comes to digital,
so maybe I can prototype MMY in virtual reality,

00:24:59.600 --> 00:25:05.400
so I can basically maybe even
compose on MMY in virtual reality.

00:25:05.700 --> 00:25:10.199
Before we actually order
the parts to be manufactured.

00:25:10.200 --> 00:25:12.700
We should make a whole like, VR game,

00:25:12.701 --> 00:25:17.501
where the purpose is to, like,
never finish the Marble Machine X.

00:25:17.502 --> 00:25:19.601
Like, the Marble Machine X is in the corner.

00:25:19.602 --> 00:25:24.499
It builds by itself and you have to
procrastinate, so it doesn't get finished.

00:25:24.500 --> 00:25:29.400 line:1
Tim Keller has his own youtube channel, so be sure
to check out Tim Keller's youtube channel.

00:25:29.401 --> 00:25:32.999 line:1
And if you wanna connect with
any of these wonderful people,

00:25:33.000 --> 00:25:36.999 line:1
just head over to the Wintergatan
discord server. They're right there.

00:25:37.000 --> 00:25:41.099
We just launched a new playlist
on the Wintergatan youtube channel

00:25:41.100 --> 00:25:44.400
with all the videos from
Marble Machine X collaborators.

00:25:44.401 --> 00:25:47.099
I don't want to miss a single video on this playlist.

00:25:47.100 --> 00:25:51.200
So if you know a video that
should be in this playlist and it's not,

00:25:51.201 --> 00:25:55.199
please contact Joanna from
Team Wintergatan on discord server.

00:25:55.200 --> 00:25:58.300
And Joanna: Welcome to the team, by the way!
[clicks tongue]

00:25:58.301 --> 00:26:01.999
A while back we asked for Wintergatan
community submissions.

00:26:02.000 --> 00:26:05.400
And we showed some of them,
but then the building took over.

00:26:05.401 --> 00:26:08.101
But there is still coming a lot of cool submissions,

00:26:08.102 --> 00:26:11.902
so we thought we should take some of
our favorites, and show them here,

00:26:11.903 --> 00:26:14.272
as a tribute to the Wintergatan community.

00:26:14.400 --> 00:26:17.800 line:1
[♪ unknown/Marble Machine (chiptune cover) ♪]

00:26:38.500 --> 00:26:41.900 line:1
Simon: - Hi, Martin. I'm Simon from a tiny
youtube channel, called Applied Procrastination,

00:26:41.901 --> 00:26:46.101 line:1
where instead of doing stuff we should have been doing,
we've been using our last couple of years

00:26:46.102 --> 00:26:49.749 line:1
building an overly complicated art project.
And we love what you do.

00:26:49.750 --> 00:26:52.750 line:1
We never even considered making
youtube videos until we saw yours.

00:26:52.751 --> 00:26:56.399 line:1
So we followed your formula and made an
open build vlog where people can follow along,

00:26:56.400 --> 00:26:59.099 line:1
and even replicate our machine for free if they want to.

00:26:59.100 --> 00:27:03.600 line:1
So this is our tribute to you and to overly
complex machines. Thanks for being an inspiration!

00:27:04.400 --> 00:27:09.400 line:1
[♪ unknown/Marble Machine (chiptune cover) ♪]

00:27:32.500 --> 00:27:34.800
[music ends]

00:27:34.801 --> 00:27:41.401
Martin: I have no clue how they made
that black fluid go into letters... [chuckles]

00:27:41.402 --> 00:27:43.502
Applied procrastination,

00:27:43.503 --> 00:27:48.999
you have the best youtube channel name
of whole youtube. Check them out.

00:27:49.000 --> 00:27:52.500
Thanks everyone for sending in
the community submissions.

00:27:52.501 --> 00:27:56.001
I can't wait until this machine is done,

00:27:56.002 --> 00:28:00.802
and I can kind of come out of this dark hole,

00:28:00.803 --> 00:28:03.699
and meet all of you on the world tour.

00:28:03.700 --> 00:28:08.300
During the spring, I was...
literally thinking of giving up

00:28:08.300 --> 00:28:12.700
and like, giving up is absolutely not an option.

00:28:12.701 --> 00:28:17.601
I believe in this project.
Even more after seeing all this.

00:28:17.602 --> 00:28:20.002
Thank you, everyone!
[poosh]

00:28:20.700 --> 00:28:24.100
Tim: I've been waiting just to
talk to you for so long, you know,

00:28:24.101 --> 00:28:27.101
take my revenge on you for calling me...
calling me a huge liar.

00:28:27.102 --> 00:28:29.802
I haven't forgotten that, by the way.
Hannes: Oh, no! [chuckles]

00:28:29.803 --> 00:28:32.303
Martin: You know what Tim Keller
told me the other day?

00:28:32.304 --> 00:28:36.003
He said that the Marble Machine X
is not sunken cost fallacy.

00:28:36.004 --> 00:28:38.304
I was so happy to hear.
[chuckles]

00:28:38.305 --> 00:28:40.505
Hannes: Okay, he's a good liar, though...

00:28:41.300 --> 00:28:44.800
Tim: But, what you missed is
that what I actually told him,

00:28:44.801 --> 00:28:47.001
was that the Marble Machine
is not a sunk cost.

00:28:47.002 --> 00:28:49.200
It's an asset. Right?

00:28:49.201 --> 00:28:54.199
And... it is, it's a brand, it's all these other things.

00:28:54.200 --> 00:28:57.200
But it's... it's an asset like a
racehorse is an asset, right?

00:28:57.201 --> 00:29:00.500
It only makes money when you ride it.
You have to feed it.

00:29:01.700 --> 00:29:05.999
So it's the kind of asset that you don't own.
It owns you. [lauging]

00:29:06.000 --> 00:29:09.999
So it's actually... It's actually worse
than if it was a sunk cost.

00:29:10.000 --> 00:29:15.100
It's an asset that completely dominates your
entire life. And if you let it go, you have nothing.

00:29:15.101 --> 00:29:17.001
[laughs]

00:29:17.002 --> 00:29:21.599
So, if Martin wants cause to be...
you know, despondent and depressed

00:29:21.600 --> 00:29:24.499
about the Marble Machine X
dominating his life, it's that:

00:29:24.500 --> 00:29:26.547
It's that he cannot get off this horse.
// Subtitled by Wintergatan Writers. Join our team on discord. //