# US 2026/0287370 A1 · System

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2026/0287370 A1
- Document ID: US20260287370A1
- Status: Published application
- Kind: A1
- Application number: 19/568,790
- Filed: 2026-03-17
- Published: 2026-09-24
- Earliest claimed: 2025-03-19
- Assignees: SoftBank Group Corp.
- Applicants: SoftBank Group Corp.
- Inventors: Takaharu ISHIKAWA
- Web page: https://patentgrep.com/patent/US20260287370A1

## CPC

- G01C 21/3415: {Dynamic re-routing, e.g. recalculating the route when the user deviates from calculated route or after detecting real-time traffic data or accidents}
- G01C 21/3461: {Preferred or disfavoured areas, e.g. dangerous zones, toll or emission zones, intersections, manoeuvre types or segments such as motorways, toll roads or ferries}
- G01C 21/3492: {employing speed data or traffic data, e.g. real-time or historical (traffic control systems for road vehicles involving transmission of navigation instructions to the vehicle G08G1/0968)}
- G06N 3/0475: Generative networks
- G06F 40/40: Processing or translation of natural language (natural language analysis G06F40/20; semantic analysis G06F40/30)
- G01C: MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY (measuring liquid level G01F; radio navigation, determining distance or velocity by use of propagation effects, e.g. Doppler effects, propagation time, of radio waves, analogous arrangements using other waves G01S)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

A system includes a processor that is configured to acquire a current location of a user by using a GPS sensor, download required map information from a map database via the Internet, and generate a prompt for instructing a generative AI model to calculate an optimal route and input the prompt to the generative AI model.

## Claims

### Claim 1

A system comprising:
circuitry configured to:
receive, via a communication interface coupled to a packet-switched network, position information indicating a current position of a mobile body from a position detection device, and acquire map information including road information, obstacle information, and traffic information from a geographic information storage device via the communication interface;
generate environment state data by integrating the position information, the map information, destination information, and constraint condition information received from a terminal device, and generate a prompt sentence based on the environment state data;
operate a generative neural network model using the prompt sentence to generate route information comprising a plurality of route points and a recommended movement condition corresponding to each route point for movement of the mobile body to a destination;
verify safety and feasibility of each route point in the route information against the map information and no-go-zone information stored in a storage device, correct the route information to remove or replace infeasible route points, and calculate an estimated arrival time based on the corrected route information; and
transmit the corrected route information and the estimated arrival time to the mobile body and the terminal device via the communication interface, receive position information and state information periodically from the mobile body, calculate a deviation between the received position information and the corrected route information, and generate a re-planning prompt sentence based on the deviation to operate the generative neural network model to dynamically update the route information.

### Claim 2 (depends on 1)

The system according to claim 1, wherein the circuitry is configured to generate the environment state data by constructing a spatial data structure integrating the position information as a coordinate tuple, the map information as a graph of road segments and junction nodes, the destination information as a target coordinate, and the constraint condition information as a weighted preference vector, and encoding the spatial data structure as a serialized text representation for input to the generative neural network model.

### Claim 3 (depends on 2)

The system according to claim 2, wherein the circuitry is configured to generate the prompt sentence by assembling segments comprising the serialized environment state data, a movement objective specification derived from the destination information and constraint condition information, and a route output format specification defining required fields for route points, movement conditions, and distance values.

### Claim 4 (depends on 3)

The system according to claim 3, wherein the circuitry is configured to operate the generative neural network model by transmitting the prompt sentence to a transformer-based generative neural network model, configuring decoding parameters to enforce route point format constraints, and parsing the generated output tokens to extract structured route point records comprising coordinate values and movement condition codes.

### Claim 5 (depends on 4)

The system according to claim 4, wherein the circuitry is configured to verify the safety and feasibility of each route point by performing a spatial proximity query against the no-go-zone information stored in the storage device, checking road network connectivity between consecutive route points using the map information graph, and applying a traversability condition check for each movement condition code against road attribute data.

### Claim 6 (depends on 1)

The system according to claim 1, wherein the circuitry is configured to calculate the estimated arrival time by applying a travel time model that computes segment traversal times as a function of segment length, road type, and traffic information values, summing the segment traversal times along the corrected route, and adjusting the sum using a temporal delay factor derived from current traffic information.

### Claim 7 (depends on 6)

The system according to claim 6, wherein the circuitry is configured to update the estimated arrival time by monitoring position information received from the mobile body, computing a remaining route distance from the current position to the destination along the corrected route information, and recomputing the estimated arrival time using the travel time model applied to remaining route segments.

### Claim 8 (depends on 1)

The system according to claim 1, wherein the circuitry is configured to calculate the deviation by computing a perpendicular distance from a current position coordinate to the nearest route segment in the corrected route information, and computing a heading deviation as a difference between a current heading derived from successive position information readings and a target heading for the nearest route segment.

### Claim 9 (depends on 8)

The system according to claim 8, wherein the circuitry is configured to generate the re-planning prompt sentence by incorporating the current position, the heading deviation, the remaining destination distance, the updated map information, and a deviation correction instruction into a structured prompt format, and transmit the re-planning prompt sentence to the generative neural network model when the perpendicular distance or heading deviation exceeds a deviation threshold.

### Claim 10 (depends on 9)

The system according to claim 9, wherein the circuitry is configured to apply the safety and feasibility verification procedure to the dynamically updated route information generated by the generative neural network model, correct infeasible route points in the updated route information, and transmit the corrected updated route information and an updated estimated arrival time to the mobile body and the terminal device via the communication interface.

### Claim 11 (depends on 1)

The system according to claim 1, wherein the circuitry is configured to receive sensor data from one or more sensors of the mobile body via the communication interface, apply a hazard detection procedure to the sensor data to detect obstacle conditions or environmental hazard conditions, and when a hazard condition is detected, generate an evacuation prompt sentence incorporating the sensor data, current position, and hazard classification.

### Claim 12 (depends on 11)

The system according to claim 11, wherein the circuitry is configured to operate the generative neural network model using the evacuation prompt sentence to generate evacuation route information comprising evacuation route points and urgency-conditioned movement conditions, verify the evacuation route points against the no-go-zone information, and transmit the verified evacuation route information to the mobile body and the terminal device via the communication interface with an elevated priority flag.

### Claim 13 (depends on 11)

The system according to claim 11, wherein the circuitry is configured to apply the hazard detection procedure by extracting feature vectors from the sensor data using signal processing operations, applying a trained hazard classifier to the feature vectors to compute hazard probability scores for a set of hazard categories, and classifying a hazard condition when a hazard probability score exceeds a hazard threshold.

### Claim 14 (depends on 1)

The system according to claim 1, wherein the circuitry is configured to update the map information by receiving differential map update records from the geographic information storage device via the communication interface, applying the differential map update records to the stored map information graph by inserting, modifying, or removing road segment records and junction node records, and using the updated map information graph in subsequent route generation and verification operations.

### Claim 15 (depends on 1)

The system according to claim 1, wherein the circuitry is configured to store the route information, the corrected route information, the deviation records, and the re-planning prompt sentences as route session records in the storage device indexed by mobile body identifier, and compute a route quality metric for each session by applying a scoring function to accumulated deviation records.

### Claim 16 (depends on 15)

The system according to claim 15, wherein the circuitry is configured to use accumulated route session records as training examples for updating parameters of the generative neural network model, assigning labels to route sessions based on whether the mobile body reached the destination without exceeding a deviation threshold, and applying a policy gradient optimization procedure to update model parameters using the labeled route session records.

### Claim 17 (depends on 1)

The system according to claim 1, wherein the circuitry is configured to coordinate route information transmission to a plurality of mobile bodies by maintaining separate corrected route information and estimated arrival time records for each mobile body in the storage device, processing periodic position information and state information from each mobile body independently, and generating individual re-planning prompt sentences for each mobile body based on respective deviation calculations.

### Claim 18

A system comprising:
circuitry configured to:
receive, via a communication interface coupled to a packet-switched network, position information from a position detection device and map information including road information, obstacle information, and traffic information from a geographic information storage device;
generate environment state data by integrating position information, map information, destination information, and constraint condition information as a spatial data structure encoded as a serialized text representation, and generate a prompt sentence assembling the environment state data, a movement objective specification, and a route output format specification;
operate a transformer-based generative neural network model using the prompt sentence to generate route information comprising structured route point records with coordinate values and movement condition codes, verify each route point against no-go-zone information and road network connectivity, and correct infeasible route points;
calculate an estimated arrival time by applying a travel time model to corrected route segments using road type and traffic information values, and transmit corrected route information and estimated arrival time to the mobile body and the terminal device via the communication interface;
receive periodic position information and state information from the mobile body, compute perpendicular distance and heading deviation, and generate a re-planning prompt sentence when deviations exceed thresholds to dynamically update route information using the generative neural network model; and
receive sensor data from sensors of the mobile body, apply a hazard detection procedure using a trained hazard classifier to detect hazard conditions, and generate evacuation route information with urgency-conditioned movement conditions for transmission to the mobile body and terminal device.

### Claim 19 (depends on 18)

The system according to claim 18, wherein the circuitry is configured to store route session records in a storage device indexed by mobile body identifier, compute route quality metrics from accumulated deviation records, and use labeled route session records as training examples for updating generative neural network model parameters via a policy gradient optimization procedure.

### Claim 20

A method comprising:
receiving, via a communication interface coupled to a packet-switched network, position information indicating a current position of a mobile body from a position detection device, and acquiring map information including road information, obstacle information, and traffic information from a geographic information storage device via the communication interface;
generating environment state data by integrating the position information, the map information, destination information, and constraint condition information received from a terminal device, and generating a prompt sentence based on the environment state data;
operating a generative neural network model using the prompt sentence to generate route information comprising a plurality of route points and a recommended movement condition corresponding to each route point for movement of the mobile body to a destination;
verifying safety and feasibility of each route point in the route information against the map information and no-go-zone information stored in a storage device, correcting the route information to remove or replace infeasible route points, and calculating an estimated arrival time based on the corrected route information; and
transmitting the corrected route information and the estimated arrival time to the mobile body and the terminal device via the communication interface, receiving position information and state information periodically from the mobile body, calculating a deviation between the received position information and the corrected route information, and generating a re-planning prompt sentence based on the deviation to operate the generative neural network model to dynamically update the route information.
