AI Designed Cars And Vertical Robotic Assembly

This could revolutionize car manufacturing.  I can see this as the future of car design and production. The simple act of integrating things like air ducts in to the car structure opened my eyes to possibility's that are just amazing. 
 

I so want to play with their software for weight reduction and stress testing.

Oh yeah,  Kevin Czinger of wildly famous Coda Automotive and Miles Electric Vehicles. Well, they're gonna need to pump out a lot of those parts for their 2 OEM contracts to get some cash flow to even think about Czinger 21C vehicles. I suppose they'll sell a few hundred of those. I would be curious to know if the Divergent software has capability to choose material based on vibration durability and environmental factors such as salt spray, humidity, and temperature variations.

P.S. obviously that back T-shirt hipster doesn't know jack about manufactur-ing and assumes everything is scaleable.

Scalability is totally in question. They are also registering there current street legal cars under SB100 Kit car registration. You can see the tag in one of the videos. 

Honestly this is a company that is abuilding a car so they can get OE companies to use them for contract design work. They will never sell more then a dozen of these things. 

When F1 uses fusion360 and structural 3d printing, that's when I think there's a revolution in carmaking. But they use carbon fiber. Cuz reasons. Good ones...

That suspension looks like it was designed by an orthopedic surgeon.  As if at some point our cars are going to be grown in a tank rather than stamped and cast at a factory.

Actually, Czinger 21C production is capped at 80 units at a base price of $1.7 million

Jay_W said:

When F1 uses fusion360 and structural 3d printing, that's when I think there's a revolution in carmaking. But they use carbon fiber. Cuz reasons. Good ones...

I get your point, but F1 has non of the cost restraints even high end road car manufacturers have to deal with.  Hand laid vacuum formed, nomex cored, CF parts are never really going to be road car feasible beyond the main tub, and then not really for true mass production.  computer refined and optimized cast parts are far more likely for mass production.

We're working with a local proto / low volume casting company for some control arms and knuckles on a project right now.  In their lobby they have a 'chassis' for one of the European space satellites that makes the control arms above look like 19th century cast iron bridges.  It's amazing.

My point of this is not so much if this company can or will be profitable it is the fact that they have shown that it can be done.  I see it as a glimpse in to the future.  Imaging the possibilities.  Not just for cars but other every day appliance type items.  

And nothing will break, ASSUMING they have taken into account all the major variations of loads. If they missed anything significant and did not bother to have the AI design with a pretty hefty margin of safety there will be a some interesting results. Things like curb strikes produce huge loads and are not something that might be considered. Also, when this sort of tech becomes common enough for it to be second hand, there will be a lot of us having to build parts out of tube and sheet because we wont know how the thing is supposed to look, when it comes in an organic shape that looks pre-bent.

RustBeltSherpa said:

Actually, Czinger 21C production is capped at 80 units at a base price of $1.7 million

Capped does not mean reached. 

So a couple things here.

• One they never let anyone see the actual 3D printers they are using, its all in a super secret room according to the Top Gear video they did. If they really had some sort of amazing 3D printing tech they would be screaming it from the room tops. So likely off the shelf stuff which means we can assume that they have standard printed properties. 
  • You can absolutely road test these designs and its not hard to take into account things like curbs on a track. If you look at really early F1 they had zero idea what they were doing in comparison to todays knowledge and they built monsters just the same. 
•  
• The build cell they have is laughable unless you are going to be bringing parts into the cell manually and handing them off with fixtures. The amount of work it would take for robotic vision and interchangeable tools grips and all the rest is way beyond what a company like this can do for a complete car. Assemblies in a press video you can do. 
  • Those robotics are not on movable jibs they don;t have a turntable, they don;t have a ton of stuff you would need to do. Plus building a car for access to that sort of cell would really put you behind your main mission allowing for access. 
•    
• Design and survivability are do very different things and when you get really out there like this using parts of the chassis for AC all sorts of livability things happen. Like mold inside your chassis. Think about pumping contaminated gas/oil out of your sealed chassis if you are using it as a vessel. Let alone having no dead spots above or below your pickup points. 
•   
• There is zero thought to repair I can bet. I wonder how long something trivial like replacing the plugs would take. 

I could go on but I am not going to. I adore this sort of design and I applaud anyone trying it. But this is not BMW using compressed carbon parts on say the i3 where they have scaled it all the way up. This is a toy to drive interest in the "AI" that they have designed. If anything in software and engineering for 24 years has taught me its once you have a good idea in software someone else is going to come up and take your lunch unless you sell quickly or seriously dominate the market. Not sure they can do that. 

A feel-good sales pitch built on claims that have holes so big you could land a jumbo her in them... My Elio-sense is tingling.

In reply to wearymicrobe :

^ This, +1,000. Totally agree that capped doesn't mean reached.