Building a Real-Time, Interactive Digital Twin for Casepak
Gaussian splats have multiple applications in M&E, but lately we are seeing more industries joining the 3DGS revolution with Volinga. Casepak, a complex industrial recycling facility, required a highly accurate digital twin to support training, maintenance, and operational planning. Traditional methods fell short of the required millimetre precision and real-time usability. That’s where UFX, a 3D and digital production team, stepped in.
By combining advanced capture techniques with Gaussian Splatting and deploying through Unreal Engine, the project transformed over 200,000 images into a fully interactive 3D environment. At the core of this transformation was Volinga, bridging the gap between raw spatial data and a usable, real-time application.
The Challenge
Creating a digital twin of Casepak introduced several technical barriers:
Accuracy Requirements: Sub-millimetre precision for engineering use cases
Processing Bottlenecks: Long reconstruction times and unstable alignments
Workflow Limitations: Traditional photogrammetry pipelines could not support real-time deployment
Without a real-time rendering layer, the project risked remaining a static dataset rather than a functional tool.
Why Gaussian Splatting
Gaussian Splatting was selected for its ability to deliver:
Real-Time Rendering: No baking or heavy mesh optimization required
High Visual Fidelity: True depth, parallax, and spatial accuracy
Scalability: Continuous updates and expansion without rebuilding
However, Gaussian splats alone were not enough. They required a system capable of making them usable in production
Volinga: The Enabling Layer
Volinga played a critical role in transforming raw data into an interactive experience.
1. Real-Time Performance at Scale Volinga enabled smooth rendering of massive datasets at 60+ FPS in 4K, making real-time interaction possible even in a dense industrial environment.
2. Data Optimization & Conversion Volinga optimized the project’s massive Gaussian splat datasets by converting raw PLY files into efficient NVOL assets, significantly reducing computational overhead for smooth real-time playback. This enabled the team to manage and render terabyte-scale datasets efficiently while maintaining high-performance interactive visualization inside Unreal Engine.
3. Seamless Unreal Engine Integration Volinga’s native integration with Unreal Engine enabled immediate deployment of the Casepak environment into a fully interactive real-time experience, providing complete control over navigation, camera movement, and user interaction while seamlessly supporting UI systems, hotspots, and embedded media layers.
4. Unlocking Interactivity The Casepak facility evolved from a static walkthrough into a fully navigable real-time 3D environment, enriched with dynamic data overlays displaying equipment specifications, maintenance schedules, and operational insights directly within the immersive experience.
Pipeline Overview
Capture: XGRIDS Portal Cam + hybrid capture system
Processing: LCC transforms imagery into Gaussian splats
Volinga: Converts and optimizes data into real-time-ready assets
Deployment: Unreal Engine for interaction and visualization
Delivery: Standalone application deployable across devices
Workflow Breakthrough
A critical pivot to the XGRIDS Portal Cam introduced structured data capture, but it was Volinga that made the dataset usable.
Without Volinga:
The project would remain a heavy, unmanageable dataset
Real-time interaction would not be possible
Deployment into production workflows would fail
With Volinga:
Data became interactive, scalable, and deployable
Processing pipelines became efficient and predictable
The environment evolved into a living digital asset
Key Takeaway
Volinga is the bridge between capture and experience. It transforms complex spatial data into a usable product, enabling real-time interaction, scalability, and deployment in environments where traditional workflows fail.
“Volinga is the enabling layer that turns raw spatial data into a usable product. Without it, this project would remain a dataset, not an experience.”
– Keshava Walcott, Technical Director & Founder, UFX