The independent claim of US12669818B2 (“Remote control request system, remote control request method, and nontransitory storage medium”), a granted patent that issued in the June 30, 2026 USPTO grant drop and is assigned to Toyota Jidosha Kabushiki Kaisha, does something worth reading carefully: it writes an operational design domain into claim language as a conjunctive test. The claim is directed to a processor that decides how a self-driving car should ask a human for help, and the pivot of the whole claim is a single determination — whether the vehicle is permitted to drive itself in a given geographic location — which the claim defines by three conditions that must all be satisfied.

determine whether the autonomous driving vehicle is permitted to implement autonomous driving within the geographic location, wherein the autonomous driving vehicle is permitted to implement the autonomous driving in the geographic location when each of the following conditions are satisfied: (a) a speed limit assigned to each roadway in which the autonomous driving vehicle is traveling and/or will be traveling according to the planned trajectory within the geographic location is less than a predetermined threshold value, (b) map information is available for each roadway in which the autonomous driving vehicle is traveling and/or will be traveling according to the planned trajectory within the geographic location, and (c) the weather data indicates that there is no rain currently within and/or forecasted to be within the geographic location for the duration of time— Remote control request system, remote control request method, and nontransitory storage medium, US12669818B2

What the independent claim actually covers

Claim 1 is a system claim reciting a memory and a processor, and its limitations run in sequence: the processor determines that the car has a planned trajectory into a geographic location for a duration; determines weather data (rain or its absence) and parking data (another vehicle parked on the roadway shoulder) for that location and window; then applies the three-part permission test quoted above. The consequential limitations are the two branches that follow. On a positive determination — the car is permitted and is driving itself — the processor requests remote assistance, in which an external operator makes a determination about the shoulder-parked vehicle while the car continues to perform autonomous driving. On a negative determination, it requests remote control driving, in which the operator remotely controls the driving operations while the car continues to not perform autonomous driving. The claim is classified under the autonomous-control class G05D (including G05D 1/0061 and G05D 1/2279).

Two features of the drafting are worth flagging for scope. First, the permission test is conjunctive — the claim uses “each of the following conditions are satisfied” — so all three of the speed-limit, map-availability, and no-rain conditions must hold for the positive branch; failing any one drops the car into the negative, remote-driving branch. Conjunctive limitations narrow a claim, because an accused system must meet every element. Second, the claim is written around a concrete triggering scenario — another vehicle parked on a roadway shoulder — rather than an abstract “ambiguous situation.” That specificity is doing work: it anchors the claim in a described embodiment rather than a functional generality.

Where the dependent claims take it

The dependent claims map the flexibility of the scheme. Claims 3 and 4 recite that the permission conditions can be either predetermined or adjustable; claims 12 and 13 do the same for the geographic location’s boundaries (preset or adjustable). That predetermined/adjustable pairing is a common belt-and-suspenders technique — it seeks to cover both a fixed-ODD implementation and one whose domain is tuned over time. Claim 6 adds the safety fallback: on the negative branch, the system verifies one or more remote-control driving conditions before handing over, and is configured to stop the vehicle or change its planned trajectory if those conditions are not met. Claim 10 introduces a server connected over a network that cooperates with a vehicle-mounted request device, and claim 11 has that server fold in supplemental information to help make the positive-or-negative call — a distributed architecture split between car and cloud. Independent claims 14 (method) and 17 (non-transitory medium) mirror claim 1’s limitations, the standard system/method/medium trilogy.

Two further dependents sharpen the sensing basis. Claims 5 and 9 recite that the permission determination is made from both external information received by the vehicle and internal information the vehicle detects or acquires directly — a limitation that ties the domain test to a fusion of onboard sensing and off-board data such as maps and weather feeds, rather than to either alone. Against the state of the art, teleoperation and remote-assistance schemes are not new to autonomous-vehicle patenting; several fleet operators have disclosed operator-in-the-loop systems. What this claim set stakes out is narrower and more specific: the conjunctive, map-and-weather-and-speed permission gate as the decision rule that selects between the two remote modes, anchored to the shoulder-parked-vehicle embodiment. Whether that specificity reads as a tight, defensible position or a limited one depends on how a given system is built — but the granted scope is unambiguous about what it requires.

This grant does not sit alone in Toyota’s June 30 block, and the neighbors sharpen its landscape. Several are classified in the traffic-warning and driver-information classes G08G and B60W: US12670795B2 (“Monitoring device”) claims warning two moving objects when an obstacle interrupts their mutual detection, and US12670793B2 (“Alarm device and alarm method”) claims prioritizing among multiple alarm targets. On the perception side, US12669336B2 claims branch-road detection from the vanishing point, and US12670738B2 (Toyota Industries) claims a distance-dependent switch between upper-body and whole-body pedestrian detection. The teleoperation grant is the decision node those perception and warning claims ultimately serve. Read together as issued claims — not applications — they show Toyota holding allowed scope across the full stack of a supervised autonomous vehicle: see it, warn about it, decide whether the car can handle it, and route the request for human help accordingly.