NEW · LIXEL L3XGRIDS’ new 5 mm flagship with built-in RTK is here — Jacob Lee is taking the first Australian L3 on the road to every state and territory.Get the spec sheet →Register for the roadshow →
Film Production · Media · Spatial AI | NEW

Capture Reality.
Create Anything.

3D Gaussian Splatting turns real-world Australian locations into production-ready digital assets for film production, virtual stages, Apple Vision Pro, and spatial AI training datasets.

Book a DemoView All Industries
The Challenge

What the Industry Struggles With

Physical production costs and immersive content demand are on a collision course.

Physical Sets Are Expensive

Building, dressing, and striking physical sets for film, TV, and advertising is one of the highest production costs. Re-shoots multiply that expense.

Spatial AI Needs Training Data

Computer vision and robotics models require massive volumes of high-fidelity, geometrically accurate 3D environments for training — and synthetic data alone is not enough.

Immersive Content Demand

The rise of Apple Vision Pro, Meta Quest, and spatial computing platforms has created demand for photorealistic 3D content that traditional pipelines cannot produce efficiently.

AECT Solution

How AECT Addresses It

From real-world scan to production-ready 3D asset in hours, not weeks.

Film Production · Set & Location Digitisation

Capture entire film and TV sets as navigable 3DGS digital twins. Directors revisit, re-light, and re-shoot inside the twin without rebuilding physical sets. Showcased at NAB Show 2026 in the XGRIDS × GeniusXR virtual-production workflow.

Virtual Production & Volumetric Stages

3DGS scenes integrated with 4DGS volumetric performances (XGRIDS × UVRSE) — cinematic worlds with real-time shadow casting, dynamic lighting, and 4D actor capture. The "Infinite Stage" workflow.

Apple Vision Pro & XR Spatial Content

Photorealistic explorable 3D environments optimised for Apple Vision Pro, Meta Quest, and WebXR. Real-world Australian locations become immersive digital experiences.

3D Training Datasets for Spatial AI

Geometrically accurate, photorealistic 3D environments at scale for training computer vision, robotics simulation (NVIDIA Isaac Lab), and autonomous navigation. Synthetic data alone is not enough — pair real captures with NuRec / OpenUSD pipelines.

Virtual Location Scouting

Scan potential filming locations once and share navigable 3D models with remote production teams. Eliminate costly travel for initial recces; share via browser link, no app install required.

Heritage & Cultural Capture

Photoreal preservation of art exhibitions, galleries, and museums — captured as living 3DGS twins in Unreal Engine 5. Side-by-side indistinguishable from physical reality. Validated on Craig & Karl's Double Vision exhibition at Griffith University Art Museum.

Capability, Demonstrated

From Photoreal Sets to Simulation-Ready Worlds

Real AECT captures — photorealistic 3DGS environments, and the same data feeding robotics simulation.

Photoreal 3DGS environment — exterior & interior, annotate and review in-browser.
Scan-to-sim — a real AECT 3DGS capture loaded into NVIDIA Isaac Sim as an AI training world.
Game-engine-ready — controller-driven flythrough of a captured environment.
Technology Stack

Equipment & Software

PortalCam

3DGS photogrammetry camera for photorealistic scene capture

LCC Studio

Reconstruction and export for 3DGS, Unity, Unreal Engine

3D Gaussian Splatting

Real-time photorealistic rendering technology

Lixel K1

SLAM LiDAR for precise interior geometry capture

Ready to Digitise Your World?

See how 3DGS can transform your production pipeline or AI training data.

Book a DemoGet a Quote