General Motors is putting artificial intelligence closer to the centre of how it creates cars, not as a showroom feature, but as a development tool.
At its Forward event in New York, GM executives said AI and advanced simulation are helping shorten the path from concept to production, with some vehicle projects now taking around half as long as they once did.
For an industry where model cycles, battery technology and software demands are all moving quickly, that’s a meaningful shift.
At a glance
- General Motors says AI and advanced simulation are helping cut some new vehicle development timelines by about half.
- The company says it can compress around 100 days of testing into a single day and evaluate major systems before building a prototype.
- GM believes its approach could bring vehicle development timelines down to around two years.
Concurrent engineering changes GM’s development rhythm
GM says it has moved away from a more traditional process where design, engineering and manufacturing teams worked largely one after another.
The company’s newer concurrent engineering approach allows multiple parts of a programme to develop at the same time.
That matters because modern vehicles are no longer shaped by hardware alone.
Powertrains, chassis systems, software, thermal management and manufacturing considerations all need to be tested against each other earlier in the process.
By bringing those streams together sooner, GM says it can find problems earlier and reduce the amount of late-stage rework that can slow a vehicle programme.
Simulation moves more testing before the prototype stage
One of the biggest changes is how much GM says it can now evaluate in simulation before a physical prototype exists.
The company says it can compress around 100 days of testing into a single day.
GM also says it can evaluate powertrain, ABS, stability control and thermal systems in a simulator ahead of prototype construction.
That doesn’t remove the need for real-world validation, but it changes when engineers get useful answers.
Instead of waiting for a built vehicle to expose conflicts between systems, teams can explore more combinations earlier and narrow the development path before metal, composites or battery packs are assembled into a test car.
Aerodynamics and battery work get faster screening
Aerodynamics is one of the areas where GM says machine learning is delivering major time savings.
The company’s machine-learning tool, trained on decades of internal data, can reportedly produce results similar to computational fluid dynamics simulations in about 60 seconds.
GM says comparable CFD work previously could take up to 60 hours.
That kind of speed is especially useful in early design, where small changes to surfaces, cooling openings and underbody treatments can affect efficiency and performance.
Battery development is another area benefiting from faster screening.
GM says a programme for new LMR cells that would normally require more than 60,000 lab hours was completed in 256 hours.
For electric vehicles, that can mean faster iteration on chemistry and performance targets before a cell candidate reaches more time-intensive validation.
A two-year target for future vehicles
GM believes these tools could reduce vehicle development timelines to around two years.
That would be a significant compression for a large automaker working across combustion, hybrid and electric vehicle architectures.
The broader point is not just speed for its own sake.
Shorter development loops can help an automaker respond more quickly to regulation, consumer demand, software requirements and battery progress.
There are limits, of course, because manufacturing readiness, safety validation and durability testing still carry real-world consequences.
But GM’s message is clear enough.
AI is becoming part of the engineering floor, not just the infotainment screen, and its most important impact may be felt long before a new vehicle reaches a dealership.
