# US 2026/0065769 A1 · System

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

- Number: US 2026/0065769 A1
- Document ID: US20260065769A1
- Status: Published application
- Kind: A1
- Application number: 19/314,329
- Filed: 2025-08-29
- Published: 2026-03-05
- Earliest claimed: 2024-09-04
- Assignees: SoftBank Group Corp.
- Applicants: Softbank Group Corp
- Inventors: Izumi Asahara
- Web page: https://patentgrep.com/patent/US20260065769A1

## CPC

- G08B 21/24: Reminder alarms, e.g. anti-loss alarms
- G06V 10/82: using neural networks
- G06V 40/10: Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G08B 21/182: {Level alarms, e.g. alarms responsive to variables exceeding a threshold}
- G08B 21/245: {Reminder of hygiene compliance policies, e.g. of washing hands}
- G08B 25/014: {Alarm signalling to a central station with two-way communication, e.g. with signalling back}
- G08B: SIGNALLING SYSTEMS, e.g. PERSONAL CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING

## Abstract

The system uses a processor to receive streaming video from cameras at set intervals, preprocess the data, and analyze it with an AI model to detect dirt or pests. Detection results, including confidence scores, timestamps, and camera locations, are stored in a database. When the confidence score exceeds a threshold, the system identifies areas needing attention and sends a notification to the user's terminal. The notification is displayed in real time as a push or in-app message. The system can also receive user questions, analyze them using natural language processing, generate answers, and display them on the user interface.

## Claims

### Claim 1

A system comprising a processor that is configured to:
receive, at predetermined intervals, video data in a streaming format from one or more camera devices;
perform preprocessing on the received video data, inputs the preprocessed video data to an artificial intelligence model, and detects at least dirt or pests in the video data,
record detection results together with confidence scores, timestamps, and camera location information to a database;
extract locations requiring cleaning or user attention when the confidence score exceeds a predetermined threshold and generates an appropriate notification message,
send the generated notification message to a user terminal;
display, in real time, the notification message as a push notification or an in-app message on a user interface of the user terminal;
and receive a question from the user, analyzes the received question using a natural language processing model, generates a specific answer to the question, and displays the answer on the user interface of the user terminal.

### Claim 2 (depends on 1)

The system according to claim 1,
wherein the processor generates the notification message based on the location data and the type of detected dirt or pest.

### Claim 3 (depends on 1)

The system according to claim 1,
wherein the processor generates, in response to the received question, a cleaning procedure or tools recommendation as the specific answer.
