The Road to Realism: How CarRig and AI Analytics Are Changing the In-Studio Car Scene
The motion realism for in-studio driving scenes has always been one of Virtual Production's toughest problems to crack.
Actors juggling performance with fake steering. Background plates that don't match the action in the car. Immersion lost in the gap between what the camera sees and what the actor can’t. The result is familiar to anyone who has spent time on a studio car shoot: broken takes, mounting pressure, and a nagging sense that no matter how sophisticated the LED wall behind the vehicle, something never quite convinces.
It is a problem the industry has accepted for years as an unavoidable cost of shooting in controlled environments. Mo-Sys CarRig was built to prove that it doesn't have to be. A compact, computer-controlled motion system that bolts onto almost any vehicle, CarRig synchronises steering, tilt and rumble motion with the background plate. The principle is elegantly straightforward: instead of asking an actor to match the movements of a video they can’t see, the system takes over the physical mechanics entirely.
The desired motion profile is created in a pre-production feed-in session, using a game steering wheel to craft the precise feel of the drive. Once approved, CarRig plays that profile back repeatedly and exactly, locked to the background content via timecode, take after take. The actor holds the wheel. The system drives it.
The gains compound quickly. Continuity across takes and camera setups becomes a given rather than a negotiation. Setup and rehearsal time on the shooting day shrinks because the motion has already been designed and signed off. Crucially, the actor's attention is freed up to focus on the scene – able to present and in the moment.
That last point is easy to understate. Performance is the thing. Everything else in the Virtual Production toolkit – the LED wall, the camera tracking, the content serving platform – exists to serve the human element. CarRig joins that ecosystem by removing one of the last remaining physical distractions from the equation.
The Wider Shift: AI Enters the VP Workflow
CarRig arrives at a moment when the conversation around Virtual Production realism is increasingly turning towards AI as a tool for faster scene generation and optimisation.
For most of VP's evolution, the challenge has been technical: get the tracking accurate, get the content rendering in real time, get the LED wall bright enough and the colour science close enough that the camera doesn't see a seam. That work is largely done. The tools exist. The workflows are maturing. Studios and productions that committed early are now shooting on LED volumes as a matter of routine.
The emerging question is subtler. Given that the technology works, how do you make it work better – not in terms of raw capability, but in terms of the output that actually reaches the audience? This is where AI-driven analytics is beginning to make a meaningful difference.
The standard workflows of CarRig are to record the motion data in pre-production, for playback on the shoot day, or to input motion on the day via the browser-based UI. However, Mo-Sys is also collaborating with partners on automated, AI-based video analysis workflows.
The June issue of Production 360 explores how Mo-Sys VP technology and AI-driven video analysis are converging to push cinematic authenticity to new levels. Read it here.
