Watch the noun in the title: not "vehicle," but "autonomous machine." NVIDIA's grant US11688181B2 ("Sensor fusion for autonomous machine applications using machine learning," issued June 27, 2023; inventors Minwoo Park, Junghyun Kwon, Mehmet K. Kocamaz, Hae-Jong Seo, Berta Rodriguez Hervas, Tae Eun Choe) is drafted to span vehicles and robots alike. That generality is the strategic core for a company that wants to be the perception layer of every autonomous thing.

The mechanism fuses multiple sensor streams through a learned model rather than a hand-tuned filter, with G06T 7/292 marking multi-sensor tracking and G06V 20/588/58 marking learned recognition of road features and objects. The claim fences the idea that fusion itself — deciding how to weight and combine modalities — is something the network learns end-to-end, not something an engineer specifies.

For the control-and-perception beat, learned fusion is a meaningful camp marker: it's a bet that data, not hand-coded sensor models, should arbitrate between camera, LiDAR, and radar. NVIDIA fencing this aligns with its Isaac/Drive strategy — own the learned perception primitives that every robot and car builder will license rather than build. The broad platform language is the point: maximize the number of downstream systems that read on the claim.

From a portfolio angle, this is arms-dealer IP. NVIDIA's robotics-and-autonomy patents increasingly use "autonomous machine" framing precisely to keep claims platform-agnostic, so they read on a humanoid, a robotaxi, and a warehouse AMR equally. Read alongside its other perception grants, this is a deliberate horizontal fence around learned multi-sensor fusion.

Caveats. "Autonomous machine" in a title doesn't guarantee claim 1 is platform-agnostic — the B60W codes suggest a vehicle embodiment, and the allowed scope may carry vehicle limitations despite the framing. Learned fusion is also heavily published; the grant turns on the specific fusion-learning step. Read the independent claim for whether it actually drops the vehicle limitation.

For the file: a horizontally-framed learned-fusion grant central to NVIDIA's autonomy-platform play. Pull US11688181B2 on PatentBear, read claim 1 for the learned-fusion step and whether the platform language survives into the claim — that determines how wide the fence really is.