Could Nio’s Safety-Certified Yoke Steer-By-Wire be Safer Than Tesla’s?
Nio claims aircraft-level, three-layer redundancy for its ET9 luxury flagship steer-by-wire. Was Tesla’s as carefully developed?We love the Tesla Cybertruck’s steer-by-wire system’s functionality so much we awarded it a Best Tech award for 2025. We like how it works and love the way Tesla continually challenges the status quo. But a recent trip to China introduced a tiny sliver of doubt as to exactly what authority—if any—signed off on the safety of that technology. The Tesla system is clearly built with redundancy—there are two steering motors fed by separate wiring, etc. But the Chinese EV automaker Nio claims that its steer-by-wire system, now in use in its Nio ET9 flaship, is the first in production to have undergone a full safety certification. The company collaborated with other manufacturers to develop a triple-layer redundancy system, one informed by aviation-industry best-practices, that then passed a Chinese government certification program.
First Level Redundancy
At this level, all computer controls, communication, and electromechanical controls are duplicated. This much everybody does. Nio also uses separate electrical architecture zones to control the twin communication and power paths, with an independent monitor comparing signals. In the event of a discrepancy, the monitor determines a “winner.” Nio, and many others offering steer-by wire use a single motor with multiple phases in the winding, wired to the disparate control systems—either of which has sufficient power to steer the vehicle (other systems provide for a fail-safe mechanical shaft connection). There are separate DC-DC converters, separate CAN bus connections, multiple steering torque sensors, and steering-yoke position sensors.
Second Level Redundancy
Here’s where we believe the Nio approach deviates from Tesla’s in following aviation’s lead. The separate zones and their communications networks were developed by different teams and sourced from different suppliers whenever possible. This helps minimize the chance of a common failure mode or use-case scenario crippling both systems at the same time.




