Wearable Camera Monitoring for Real-Time Abnormal Behavior Detection

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional technologies fail to monitor user safety in real time and promptly detect abnormalities or dangerous situations, leading to potential risks and delayed responses.

Innovation Solution

A system comprising a wearable camera that records and transmits video to a cloud server for real-time analysis, using generative AI to detect abnormal behavior or dangerous situations, and notifies the user's family or related parties via a smartphone application.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If conventional chatbot systems are used, then basic user interaction is maintained, but real-time safety monitoring and abnormality detection are insufficient

Engineering Contradiction:

Improveuser safety monitoringVSAvoiddetection response time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The patent replaces manual safety monitoring with an automated computer vision system using deep learning models. The analysis unit automatically processes video feeds from wearable cameras, detects abnormalities through trained neural networks, and triggers alerts without human intervention, achieving both high reliability and real-time response

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent introduces a cloud-based analysis server as an intermediary between the wearable camera and the user/family. This intermediary receives video data, performs complex abnormality detection using generative AI models, and returns results, enabling sophisticated safety monitoring without burdening the wearable device itself

Inventive Principle:

Principle #24Intermediary (Mediator)

2Reliability

If video recording and transmission is implemented, then abnormality detection capability is improved, but data transmission load and processing requirements increase

Engineering Contradiction:

Improveabnormality detection accuracyVSAvoidvideo data volume

Core Design Contradiction:

ReliabilityVSQuantity of substance

Solution Approach 1:

The patent extracts only the essential visual information needed for safety monitoring from the video stream. The analysis unit processes video feeds to identify specific abnormal patterns (falls, unusual movements, dangerous situations) without transmitting or storing the entire video dataset, reducing data volume while maintaining detection accuracy

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent implements continuous video monitoring and transmission to ensure no abnormality is missed, then applies AI analysis to the transmitted data. This partial action approach transmits all video data for comprehensive analysis, accepting the data volume burden to ensure complete coverage of potential safety incidents

Inventive Principle:

Principle #16Partial or excessive action

Data Source

PatentUS20260065771A1system

Publication Date: 2026.03.05 SOFTBANK GROUP CORP

AI summary

The system according to the embodiment comprises a recording unit, a transmission unit, an analysis unit, and a notification unit. The recording unit records video of a user wearing a wearable camera. The transmission unit transmits the video recorded by the recording unit to a cloud server. The analysis unit analyzes the video transmitted by the transmission unit and detects abnormal behavior or dangerous situations. The notification unit notifies the user's family or related parties of abnormalities or dangerous situations detected by the analysis unit.