DETROIT, April 14, 2026 (GLOBE NEWSWIRE) — At SAE World Congress 2026, Luminary, the Physics AI company, today announced SHIFT-Crash, the first Physics AI model that predicts full-vehicle crash response, including deformation and stress fields, in seconds instead of hours. Built on a dataset of 5,000 crash simulations based on the 2010 Toyota Yaris, SHIFT-Crash is the first learned surrogate to capture crash dynamics across an entire vehicle and a multi-dimensional parametric design space, generating spatially resolved stress field predictions directly from vehicle design parameters.
Unlike traditional finite element method (FEM) simulation, which must be run from scratch for every new design variant, SHIFT-Crash is a reusable AI model that retains and improves on crash physics knowledge across vehicle programs.
The Crash Simulation Bottleneck
Crash simulation is one of the most computational expensive stages of vehicle development. A single NHTSA NCAP 56 km/hr full-vehicle frontal crash simulation typically takes 10 to 12 hours on HPC clusters. A typical vehicle program requires thousands of such simulations across load cases and design variants. Critically, traditional FEM simulation has no memory. Every vehicle program must start from zero with a new mesh and solver run because design knowledge from previous platforms cannot be transferred. As a result, detailed crashworthiness analysis typically occurs late in development, when design changes are minimal and far more expensive to implement.
From Days to Seconds: Expanding Crash Design Exploration
The automotive industry is increasingly adopting virtual validation as part of the crashworthiness development process. In February 2026, the BMW Group became the first carmaker to have virtual crash simulations officially recognized as equivalent to physical tests in Germany. SHIFT-Crash introduces a new capability within this workflow. By predicting full-vehicle crash response in seconds, the model allows engineers to evaluate far more design variations within a single development cycle. A major automotive OEM confirmed that this speed would enable engineers to explore design spaces previously inaccessible within program timelines, accelerating crashworthiness optimization earlier in vehicle developments.
The business impact is straightforward. OEMs that can screen designs against NHTSA NCAP, Euro NCAP 2026, and IIHS protocols in seconds rather than days can lock designs earlier, move vehicles into production sooner, and reach showrooms ahead of competitors still waiting on HPC simulation queues. For high-volume programs, even a one-to-two month acceleration in market entry can translate into significant revenue upside, in addition to the cost savings Physics AI already delivers.
A Model That Gets Smarter with Every Program
SHIFT-Crash uses transfer learning to carry crash physics knowledge forward from one vehicle program to the next, and across vehicle classes. As the model is applied to additional programs, it continuously builds on what it has already learned about structural crash behavior.
For an initial SUV program, an OEM may train SHIFT-Crash using approximately 5,000 crash simulations or existing historical simulation data. As the model accumulates knowledge from successive programs, the number of new simulations required to evaluate future designs can drop dramatically. By the third SUV program, fewer than 300 additional simulations may be needed because much of the underlying crash physics has already been captured.
This compounding knowledge enables engineers to evaluate far more design variations within a single development cycle, accelerating crashworthiness optimization and reducing the time required to validate new vehicle programs.
When the OEM moves to a new vehicle class, partial transfer still applies. Structural crash physics learned from the SUV programs means that the first sedan program may require roughly 600 simulations instead of starting again at 5,000.
Key Benefits of AI for Crash Testing
- Compounding Cost Savings: Transfer learning capabilities allow knowledge to carry across vehicle programs, significantly reducing the number of expensive simulations required for future designs.
- Rapid Design Optimization: Engineers can quickly explore design changes across both vehicle geometry, such as crush zone structure or cabin dimensions, and material parameters, such as rail thickness, much earlier in the development cycle.
- Trustworthy Downstream Analysis: With workflow-ready predictions demonstrating less than 3 percent Root Mean Square Error (RMSE), engineers can confidently perform downstream safety analysis and derive key crash metrics.
“Crash analysis today is a very expensive and time-consuming computational mechanics problem. With SHIFT-Crash, we offer crash engineers the ability to explore richer design spaces than ever before, compress time to solution and drastically reduce costs,” said Suds Menon, Chief Product Officer at Luminary. “With a Physics AI Crash model, each new vehicle design program reduces the number of simulations required, makes the model smarter, and delivers higher accuracy results in a fraction of time, compressing design cycles.”
About Luminary
Luminary is a Physics AI platform for rapid design iteration, design exploration and optimization. Customers span industries from aerospace and defense, automotive, to leading industrial manufacturing companies, including Otto Aerospace, Sceye, and Welbilt. For more information, visit http://www.luminary.ai.
A photo accompanying this announcement is available at https://www.globenewswire.com/NewsRoom/AttachmentNg/977c2061-d09d-4de6-a304-b0c6f28e4299
