» Mercedes EQC Scanner (Insights)
This site relies heavily on Javascript. You should enable it if you want the full experience. Learn more.

Mercedes EQC Scanner (Insights)

credits wirmachenbunt / Atelier Markgraph

This is an attempt to show a little bit more of a project, than just a video. There are some interesting bits and pieces, relying on brilliant contributions and sometimes overlooked cookies.

But first, let's have the video anyway.

The EQC-Scanner is sort of a augmented, kinetic installation. One could argue with the term augmented here but it certainly adds information to the "real" layer. The whole thing is controlled by a touch screen, allowing you to pick topics or move the screens with your fingertip. It is all and all car technology communication, but in a playful package.

Ok,let's have a look inside.

Fractal Lightning

This is based on some articles like this LINK A pretty neat recursive routine to learn what recursive is. I used c# but i bet this is easy in VL, anyone ?

The plugin was used for the battery scene. Thank you captain obvious :)

Binary Operators

Usually, when you are in RS232 or some fancy protocol land, you have to decode and encode values efficiently. Like encode high values with little use of digits. I always come back to jens.a.e'sBitWiseOps, this is one of the overlooked contributions.
For this project, the plugins were used to encode the engine controller messages.

Analysis

The image above shows some sensor recordings of the screen movement. Creating this data viz in vvvv was quick and revealed a physical feedback loop. The violet curve shows the frame difference of the screen position. And while the real screen movement actually looked kind of smooth, the sensor data showed some heavy regular spikes. Obviously, the engines did some regular overshooting. Not a big problem and solvable on the hardware side. Interesting how data visualization can help to track down problems.

Vectorraygen + DX11.Particles

Controlling Particles in a meaningful way can be painful. Using vectorfields can bring some structure into the chaos. The tool Vectorraygen helps generating the vectors the easy way. It even has a node-based enviroment to drop some organic variation into your fields. And btw. the devs are very friendly.https://jangafx.com/

The tool was used to create the key visual, 500k floating particles along the car exterior.
DX11.Particles

Boygrouping

Sure, this is not a big secret, it's one of vvvv's selling points. But i have to say, it just works. Bringing together 3 machines was actually fun. Hopefully VL preserves this killer feature.
Boygrouping

Automata UI

This was pretty much the first time i really used my own plugin and it was surprisingly helpful. :) It consists of a initialization tree for the engines & sensors and an abstract transition model of the visual software part. This is an attempt to leave content out of the statemachine but rather use the state TOPIC for every topic in use. It might be harder to read and it doesn't allow you jump to a specific content, but it makes your software expandable (without touching Automata again).

Not sure if this is a best practise for everything, let's see how this works out in the future.
Automata UI

Content Management System

There is always the dream of a fully, totally dynamic vvvv project. Our ruby-on-rails based web tool helps to manage all texts and images. It even renders the texts as images, freeing vvvv from rendering crappy text but rather display smooth pre-rendered images. Most of the vvvv project is depended to the CMS but of course, there are some bits which are always hard-wired, like 3D stuff.
CMS

I hope you find this "article" informative, any questions, comments ?

Cheers

Chris
wirmachenbunt

u7angel, Monday, Nov 5th 2018 Digg | Tweet | Delicious 8 comments  
tonfilm 06/11/2018 - 03:52

thanks @u7angel, really useful and interesting to see how you use the tools around vvvv to create such a complex project.

CeeYaa 06/11/2018 - 09:28

short simple detailed description of a visually nice and technical complex project - amazing, thanks - love this kind of description

hrovac 06/11/2018 - 09:30

thank you Chris, working in the Daimler-UI-Design i find this even more interesting ;)
i will show this in our next open discussion.
also, i will have a look at the vectorfiled tool..

sebescudie 06/11/2018 - 10:38

nice clean and interesting insights, thanks u7!
makes me realize I totally overlooked this CMS thingy, will check it out :)

colorsound 06/11/2018 - 10:45

very cool , congrats and thanks for sharing descriptions.

Noir 06/11/2018 - 20:42

Cool stuff and nice description too

m4d 07/11/2018 - 17:50

Great project breakdown. Thanks for giving us a peek behind the curtain!

u7angel 08/11/2018 - 12:12

happy to hear such a breakdown is of interest. i added one minor bit.

  • 1

anonymous user login

Shoutbox

~22h ago

tonfilm: @beyon you can also use F# together with #vl, see: #fsharp #dotnet #visualprogramming https://github.com/vvvv/VL.DemoLib/tree/master/src/VL.FSharpDemoLib

~1d ago

beyon: tonfilm: Ok,currently using F# though but maybe I should look into sharpdx vs xenko math

~2d ago

tonfilm: @beyon #vl also fills these gaps with huge range of collision detection math from #sharpdx which is basically the same as #xenko's

~2d ago

tonfilm: vvvv #audio pack update: better buffer reader and fixed #vl template #visualprogramming #creativecoding #dotnet vvvv.audio-pack-alpha

~2d ago

beyon: microdee: Good, I found Xenko.Core.Mathematics filled many gaps in SlimDX ray/point/plane/bbox functions

~2d ago

microdee: "if you want to make a custom project only using Xenko graphics API without the full Xenko engine, you can!" https://xenko.com/blog/feature-spotlight-xenko-3-1-and-nuget/

~3d ago

m4d: @circuitb: looking good :)

~4d ago

joreg: berlin, london, paris, who's next? thanks for organizing this @sebescudie

~5d ago

sebescudie: in Paris and into #vvvv ? Let's have a bier next Wednesday ! https://bit.ly/2z4BnKQ