Objective: Vizualize a simulated car crash in VR, at real scale, and compare it with the scanned result of a real crash
Work: Simulation & scanned meshes processing, rendering optimization, UI
Our client Actoat Engineering is a leader in simulating car crashes with an incredible accuracy, and was considering VR as a tool to improve quality control and simplify the data extraction for the end user.
For this, we processed one of their simulation, as well as the scanned result of the real, same crash. The end goal was to offer the ability to visualize the simulation at scale, at different speeds, and be able to save car parts at different timesteps for a closer look and comparison. It was also mandatory to be able to compare the scanned mesh and the simulation to account for the accuracy.
Processing the meshes was a real challenge as the raw data is beyond what any VR engine could display in real-time. Lot of work went into building a pipeline to process the meshes and made them VR-friendly, while retaining the layers and details for proper inspection of the models inside the app.
UI was another critical aspect of the app, to allow an intuitive experience while offering maximum features.
A new version is currently undergoing development, with new features (scaling, removing components, etc...) and an automated pipeline, allowing the client to process its own meshes and store them inside the app.
Simulation control by intuitive UI or by controller's buttons.
Model can be detached to grab/inspect each component individually.
Immediate comparison between the simulation and the 3D scan.
User can stand inside the mesh for closer inspection, or can use a side-by-side view with hotspots.