ROADSTER FLIGHT LAB
EXPERIMENTAL VEHICLE DYNAMICS / 001
NEVADA / PROVING GROUND

Runway

READY WHEN YOU ARE

A different kind of drive.

Start on the runway, or choose a test setup.
Keep some gas for the landing.

WASD drive Space lift
STANDBY00:00.0
N
LIVE TELEMETRY
Altitude AGL0.0m
Vertical speed0.0m/s
Load factor1.0g
Suction force0.0kN
Lift / cushion0.0 / 0.0 kN
TARMAC4 CONTACTS
Surface seal0%
000
THRUSTER GAS100%
200 kN·s available
BATTERY100%
ASSISTED FLIGHT
LAST TOUCHDOWN— m/sAwaiting landing
ENGINEERING / PARAMETERS

Tune the experiment.

These are adjustable simulation assumptions. Changing hardware resets the test so comparisons begin in the same state.

Fan cushion is a hypothetical near-ground pressure effect. It uses battery energy and a spool delay. It is not a verified Roadster capability.

PILOT BRIEFING

You're in control.

W / SAccelerate / reverseA / DSteer on groundSpaceLift thrustersShiftRear boost thrustersQ / ERoll left / rightI / KPitch forward / backJ / LYaw left / rightBWheel brakesF HToggle suction / flight computerP RPause / resetC VCamera / force vectors

Assisted flight levels the car and brakes a descent using real simulated gas. In manual mode you control attitude and landing thrust. It cannot rescue you with an empty tank. Drag the scene to orbit in Orbit camera; scroll to zoom.

RESEARCH / MODEL V1

Physics, with the assumptions visible.

This independent simulator explores the fan-car and cold-gas concepts described in the October 2026 Roadster reveal. It is not affiliated with Tesla or a validated engineering model of its car.

What is modeled

Six-degree-of-freedom rigid-body motion, four spring/damper tire contacts, load-limited tire friction, nearby-surface suction, aerodynamic drag, rear boost, lift and attitude impulses. Gravity remains world-down. Fan seal drops as the vehicle separates from the surface.

Gas and the flight computer

The tank is represented by usable total impulse (kN·s), not guessed tank pressure. Every propulsion and attitude command spends that shared budget. Nozzle-flow, temperature, pressure regulation, changing propellant mass and real control allocation are simplified. Assisted landing uses the same force and gas limits as manual input.

The reverse-fan experiment

No verified source found says reversed fans softened the reveal landing. This optional model assumes suitable ducts and skirts can create an air cushion close to a surface. Force falls off with clearance and takes time to build. Its strength is adjustable independently; that is a design experiment, not an assertion that Tesla's fans can do it.

Starting assumptions

1,900 kg total mass; 1 MW drive power; 2 body-weights of suction; 60 kN rear thrust; 42 kN lift; 200 kN·s total impulse. These intentionally do not combine all published best-case claims. Aero and tire coefficients are simplified. The battery model assumes 115 kWh for energy accounting; this is not a confirmed production battery capacity.

How to compare landings

Select Landing study. Run once with the flight computer and cushion off, once with only the cushion, and once with only the flight computer. Reset returns the same height and full resources. Compare touchdown speed, gas remaining and exported telemetry. Lower speed is a softer landing; the load-factor reading includes approximate suspension forces.

Known limits

Surface contacts use rays and simplified hull contact. Tires, suspension, aerodynamics and collisions are approximations; narrow edges and extreme speeds can expose them. No fluid simulation, real Tesla flight controller, production certification, water physics or damage mechanics. The facility is a testing ground, not a real location map.

Research sources

Later: the game

Hudson's backlog: canyon and bridge jumps, ponds, commute shortcuts, a departing car ferry with a time and gas budget, a GTA-style world, manual and automated controls, and possible Stripe upgrades. No payments are connected in this simulator.