If you mean the side view at 40-45 seconds, those wheels are still on the ground, still supporting substantial weight. But their individual suspensions are raised up quite a bit compared to the section right under the saddle supporting the vessel. Every set of wheels is maneuverable. If they're just moving the trailers, they'll lift a lot of them. (Although they'll stack them up to move them more efficiently.) With a heavy load, they're all on the ground. I don't know what the control system looks like, but I imagine it's a fairly complex block of code, reacting in real time to terrain changes, side forces from wind and turning, pressure loss in individual tires, etc. Getting back on topic (Sorry for derailing this) To model a SPMT (self-propelled modular trailer) in BeamNG, you could get away with not modeling every wheel set completely. You could have a lot of dummy wheels modeled into the body of the trailer (like all of the inside wheels). The ground doesn't deflect, so you wouldn't have the real-world concern of destroying the road. A couple dozen active tires along the endges would be less resource-intensive because you could use a rigid, non-deforming material to mimic doing the job of a hundred or more. The inner tires could be a non-colliding material to avoid complications with uneven road surfaces. Tires wouldn't be switchable, and pressure wouldn't be changeable. You could use maybe eight suspension units per trailer, with straight vertical travel, heavily damped. You wouldn't even need to see them from the outside, just model one giant spring and shock for each unit. To make them modular, you'd need rigid couplers on all four sides. You could have power units to mount front or back. In @PIISIGMA's pic, the power units are high-mounted at the front. There are also units that are no higher than the trailer, and nearly touch the ground, mounted front or back. Mounting a load on a single trailer would be easy: 2 to 4 saddles coupled using nodes. No relative movement necessary. For something like the Heartland video, you'd need to model the base plate and swivel joint below the saddle in the middle of each trailer group. (Each trailer in that project is actually four trailer modules.) Truck modifications To modify the truck, you'd add a weight on the fifth-wheel, and front and rear tow bars coupled to the trailer or to another truck. (Notice the tandem tractor arrangement in the video.) You'd use the biggest engine available, and the model already has snow chains for winter moves like we do here in Alberta. (Loads like this have to be moved in winter when the ground is frozen and the bridges are stiffer. So even though the Big One will be done in a couple of months, it will sit in our yard until next winter.) For multi-tractor moves, you'd also need a way to control multiple trucks at once (or have really good coordination with a multi-player mod). All in all, it would be a hell of a project. As a complete programming noob, I wouldn't be much help except for providing limited input on the real-world version. I don't work with the trailers themselves; I manage the health & safety of the company that builds the pressure vessels. But I've got a mechanical design diploma, so I understand the mechanics of it all. Extra trivia Just FYI, here are some specs of the bigger version of the vessel in the Heartland video, the one we're building now: Length: 340 feet (Will be like a 35-story building when lifted upright) Diameter: 28 feet Empty weight: 850 tons (Just the vessel. Total weight with trailers will be close to 2000 tons.) Internal volume: 5,450,000 liters (1,440,000 US gallons) ...and we have to fill it up with water to hydro test it. It takes about a year to build something like this. But if you think that's impressive, I talked to one of the Mammoet guys before they moved the vessel in the video, and he told me that the previous year, in Russia, they moved a load ten times that size.