Real-Time Automotive Marketing with NVIDIA Omniverse
Modular Application Building with Omniverse and APIs
Katana needed the flexibility to build a solution that suited their needs, aligned with their skillset, and interoperated easily with their existing pipeline. Omniverse’s modular components allowed them to integrate only what they needed from the platform. For example, the COATcreate high-fidelity viewport provides photoreal, ray-traced rendering, powered by the Omniverse RTX Renderer. The overall experience and user-friendly interface is delivered over the WebGL API, and pixels are streamed from the cloud.
COATcreate accesses CAD data, represented as USD, allowing a user to visualize and change scene data in real time, and manage a set of preconfigured variant sets such as trim, colors, and wheel options. Katana also built simple animation features such as animated cameras, and interactive car components including doors and the hood. This enables the user to easily create stills or animations for any project.
"COAT and USD unlock the ability to create high-quality, photorealistic content, enabling designers and art directors to create what they want in real time," said Cappy Childs, chief operating officer at Katana Studio. “Previously, you spend all this time preparing data for one shot, and then throw it away. Now we can repurpose it later for other shots. It’s like a paradigm shift for artists and clients because they're creating something with longevity versus single use.”
The Katana team developed technology to further leverage OpenUSD and Omniverse integration, named Virtual Variant Sets. These simplify how users access and toggle through multiple variants of different assets, such as the car or environment options that have been created during the data prep stage.
“We also capture metadata in the USD file and incorporate it into rendered stills, allowing users to share a live link,” said Damian Fulman, global director of technology at Katana Studio. “This way, they can easily return to the same state or reload the image in the future, even if there are changes to the scene or the car.”
Nissan Enhances Product Validation and Data Preparation
Leading automotive manufacturer Nissan is leveraging COATcreate for multiple purposes, including product validation and marketing asset creation. The team at Katana prepared and integrated Nissan's data into COAT, enabling Nissans users to seamlessly interact and test configurations of the new Nissan Z.
By leveraging COATcreate, Nissan can quickly and efficiently validate configurations of the Z, ensuring they meet the desired specifications and standards. This streamlined process allows for faster iterations and improvements, ultimately enhancing the overall quality of Nissan's vehicles.
In terms of time savings, the traditional production pipeline—which involves multiple rounds of the clients interacting with the agency, and the agency communicating with the studio—would typically take about three weeks to complete a project. COATcreate allows users to execute that same process in hours.
“The Nissan Virtual Garage has begun working with an OpenUSD workflow, and we expect to realize a 30-40% time savings in the first year. OpenUSD is a paradigm shift for automotive digital twins, and we expect to radically change not just our own workflow, but how CG assets and configurators will be created in the very near future.” said Chad Taylor, head of the Nissan Virtual Garage and digital Nissan assets.
As for what’s next for Katana, the team is looking forward to seeing how auto companies will adopt OpenUSD, especially as more leading organizations join the Alliance for OpenUSD.
“As companies move away from old pipelines and start adopting new pipelines, they’re going to use USD like this,” said Fulmer. “This is the best of all the worlds coming together, where we can build on the shoulders of giants to be able to make things like this that we couldn’t do on our own. We can write new tools to manage datasets in a much more efficient way, which results in a tenfold speed increase in production pipelines.”
source: link
