The CAVERN is a CAVE (Cave Automatic Virtual Environment) system that is uniquely curved around a 270-degree panoramic, 20 foot long screen. This creates a large open space and a continuous surface on which to project images, which, combined with the CAVERN’s stereoscopic 3d technology, can create a uniquely freeing VR environment.
When the CAVERN was developed at CMU’s Entertainment Technology Center, it was always intended to be used for multiplayer VR experiences. Unlike VR headsets, the open space of the CAVERN allows players to look at and interact with each other physically, and move around the space without having a heavy piece of equipment on their faces.
However, while the CAVERN already has the capability to track the position of one player in its space, and render the world from their perspective, it is limited to just that one player. Any additional players would need to closely herd around them in order to try and match their perspective.
Over the summer of 2025, Steve Audia, the creator of the CAVERN, had an idea about a method that could potentially solve this problem. By rendering objects that are projected inside the space of the CAVERN multiple times, once for each player, they would all be able to see the same object from the correct perspective. This would enable new interactions in the CAVERN. Instead of following one “main character”, players would be able to hand off objects to each other, face to face, or look at the same object from different angles and communicate information with each other live.
Realizing this idea is the goal of Team Torchbearer. Building on previous work in the CAVERN, particularly Spelunx’s CAVERN toolkit, our first step is to implement this technique into the rendering of objects on the inside of the CAVERN space. Our second step will be to experiment with the technique, find ways to use it, and see what its limits are. Our third and final step will be to develop a game that uses it, and demonstrates the CAVERN’s capabilities as a truly multiplayer VR environment.