Nathan Baker, VP of rapid engineering design in the central analysis team at Arrival, believes there will come a time when real prototypes are not needed.
Traditionally, the automotive industry has been reliant on physical testing because it was forced to be – simulation tools and computational power weren’t sufficient. More recently, simulation capabilities have developed tremendously.
The fidelity of simulation models is increasing, and the underlying physics modeling is improving. Our understanding of how to use simulation is also getting better, and computational power is almost at
the point where it’s not a limiting factor, especially with high- performance cloud computing.
However, most people still suggest that testing will always be required. In the past, people [working]in testing would say, “Testing is the truth.” In the near term that might well be the case, but if you asked a simulation expert about this today, they likely would say that, “Simulation is the truth.” The reality is somewhere in between. Nevertheless, we need to challenge the testing mindset and push into simulation heavily because when a model is well correlated, it’s faster and more cost-efficient and it can be used to evaluate a multitude of conditions in quick succession.
Simulation enables us to develop products rapidly and effectively de-risk a successful launch. It gives freedom to explore concepts at the earliest stages of design that you just can’t do with physical prototypes. Ultimately, it enables a more mature product that achieves vehicle attribute targets sooner.
An example of simulation-based product development is the use of the driver-in-the-loop simulator, which
we use across all of our platforms at both conceptual architectural development and detailed design phases. It has been incredibly valuable to us as a company, because while we’ve been waiting for our prototype vehicles to be manufactured, we’ve been able to perform vehicle dynamics development to provide guidance to engineering on the direction we want to take. The ability to shape the vehicle definition at all stages of our programs means a more robust product and lower development costs.
In addition, there are some things that you can’t measure in a test without affecting the test itself. For example, the addition of instrumentation to an assembly means that the instrumentation itself may impact what’s being measured. This is where simulation and analysis comes into its own, as through a correlated model you have the opportunity to understand and compute what you couldn’t otherwise measure through testing.
Red-Cat’s dream is 100% digital vehicle sign-off. Although that may be a moonshot and an alien concept to many, that goal in itself will lead to a more sustainable future. Building prototypes consumes energy and materials, which could be partly offset with a simulation-driven approach. And the more simulation and virtual development we can do, the fewer physical prototypes we need to build.
However, to reach this digital sign-off goal, we need to gain confidence in and a better appreciation of the models being created, which means we’re having to do a lot of physical testing. At the moment, testing is important because it’s what enables us to get better quality inputs into our models. The correlation of our models in these early stages builds confidence so we will need less and less testing in the future.
At the vehicle level, we are digitally modeling all of our products so we will be less reliant on prototypes in the future. In the near term, I see that we will still need some prototype vehicles but we will tend to more strategically do component and assembly-level testing for correlation and early feedback on design directions.