NEC's 2020 Motion-Planning Grant Plans for Objects That Move — And the CPC Says It's a Vehicle

"Autonomous moving objects" is the kind of phrase patent drafters use when they want a claim to travel. NEC's grant US10802494B2 ("Method for motion planning for autonomous moving objects," issued October 13, 2020; inventors Francesco Alesiani, Chairit Wuthishuwong) plans motion for a self-moving system that has to account for other things in its environment also moving. The generality of the title is a scope play; the CPC is where the real target shows.

The mechanism is dynamic motion planning: instead of planning a path through a static snapshot, the method reasons about how the scene will evolve and commits to a trajectory that stays feasible as obstacles move. The G05D 1/02xx position-control family and the G01C 21/005 navigation code mark the planning-and-navigation core, and the G05D 2201/0213 tag — the standard CPC stamp for land vehicles — quietly anchors the broad-sounding claim to ground autonomy.

For the control beat, the giveaway is the presence of G05D 1/0274 (vision-based position control) alongside the planning codes. It signals a claim that closes the loop from perception to plan rather than assuming a clean map handed down from somewhere else. Planning claims that ingest live perception age better than ones that assume a pre-built world, because the hard part of real autonomy is exactly that the world won't hold still.

NEC is not a name people associate with the AV wars, which is part of what makes the filing interesting from a portfolio angle. A horizontally-scoped motion-planning claim from a large diversified electronics company is the kind of asset that gets asserted or licensed across industries rather than weaponized against a single robotaxi rival. The breadth in the title supports that posture.

Caveats, stated plainly. A title that says "autonomous moving objects" does not guarantee claim 1 is platform-agnostic — the land-vehicle CPC tag suggests the examined embodiment was a ground vehicle, and the allowed scope may carry vehicle-flavored limitations. And motion planning amid moving obstacles is a deep prior-art field; the patentable core is a specific computational step, not the concept.

For the file: treat this as a mid-breadth, perception-coupled motion-planning grant from a non-obvious assignee. Pull US10802494B2 on PatentBear and read whether claim 1 actually recites the prediction-of-other-objects step — that is the limitation that separates this from ordinary static path planning.