ORLANDO – A U.S.-Czech startup introduced a new mixed reality headset that the company says will sharply reduce latency issues.
Vrgineers’ XTAL 3 Ceiling and Visibility Unlimited, or CAVU Mixed Reality Headset, is primarily intended pilot and aviation training, and will be sent to the U.S. Navy and the Ukrainian Air Force in the mid-2024 for evaluation, a company official said.
“We are supplying and improving these headsets according to requirements from the Naval Warfare Center, because we are under contract with them and they are providing us feedback,” said Marek Polcak, the CEO and co-founder of Vrgineers.
The company unveiled the new headset at the National Training and Simulation Association’s Interservice/Industry Training, Simulation and Education Conference on Nov. 30.
Along with the Naval Warfare Center, Lockheed Martin and the Ukrainian Air Force will also be receiving the headsets for test and evaluation, he added.
“The problem is that with existing headsets, we have reached their limitations. We are unable to upgrade anything anymore,” Polcak said. “The limitation is the system, the architecture. To improve mixed reality, we have to come up with a new design, a new platform, new architecture, and this is what we did.”
The XTAL 3 CAVU Mixed Reality Headset utilizes a completely different and streamlined architecture compared to standard mixed reality headsets, providing users with advanced visual quality, latency reduction and a lighter, more comfortable fit, he said.
The headset, developed by Vrgineers — along with with graphical chip makers AMD and NVIDIA — is made of “precision-engineered” carbon, he said, which allows for the lighter weight.
Higher-quality chips ensure reduced latency and limited compression, resulting in better-quality images, also reducing noise and blur, he said.
“We are simplifying the flow of the data for mixed reality, and that’s the big difference, because this allows us to utilize much higher-quality chips in the headset,” he said.
“The chips [we use are] 24 MPix with 12-bit color depth, it’s unheard of,” Polcak said. “They are closer to Hollywood cameras than to mixed reality chips, because today everyone is using chips from mobile phone cameras. We are using professional chips with much higher quality.”
The big differentiator between Vrgineers’ streamlined architecture and conventional mixed reality architecture is the FPGA processing step, which condenses the three steps of headset image processing, CPU and RAM processing and GPU image processing into one step in the architecture. This is crucial, because between each “step” in the processing phase, there is a brief but important delay of “between one to five milliseconds,” Polcak said. Less steps means less time wasted.
“With this architecture, we are able to do all of these [processing steps] in only one millisecond guaranteed. And that’s the big difference in the latency,” he said.
The headset will be embedded in Vrgineers Mixed Reality Classroom Simulators and available for purchase “mid next year” for selected partners, he added.
“It’s quite a unique architecture. And we have to be cautious, regarding our resources and where it will be deployed,” Polcak said.
Source: National Defense
