A two-class CPC is a signal to read the claim narrowly, and GM's grant US10678253B2, "Control systems, control methods and controllers for an autonomous vehicle" (issued June 9, 2020), carries just two: G05D 1/0223 (autonomous control responsive to the environment) and G01C 21/20 (navigation). That keeps the claim on the control layer — turning a plan and a position into actual vehicle motion — rather than on perception.
The enabling question is what the controller specifically does beyond 'control an autonomous vehicle,' which would be unpatentably broad. The pairing of an environment-responsive control class with a navigation class implies a claim tying navigation state to control output, and the defensible novelty has to be that specific coupling. The generic title hides the substance; the limitations carry it.
On scope, the independent claim establishes the controller-and-method combination; the dependents that specify the control law and the navigation inputs are the moat. As a granted B2, the scope survived examination, so the limitations are concrete enough to distinguish over general AV-control art. A dependent that ties the control response to a defined environmental condition is where the real protection sits.
Strategically, GM filing steadily on control-layer IP in 2020 — alongside its monitoring and rule-learning grants from the same year — paints a picture of a company protecting the full autonomous stack, not just the flashy perception parts. The verdict: enabling, narrow, and examined. Read it as a specific control-method grant, label nothing from the abstract as the invention, and find the substance in the navigation-to-control dependents.