Market Context — Why This Technology, Why Now

The global automotive and logistics sectors are undergoing a rapid transformation driven by stringent safety regulations, the proliferation of ADAS, and the imperative for operational cost reduction. Companies are seeking advanced AI solutions to manage vast amounts of vehicle data, mitigate accident risks, and optimize fleet performance. This technology offers a critical tool for meeting these demands, enabling proactive safety measures and data-driven decision-making across diverse vehicle applications.

Key Competitive Advantages
01

Significantly Reduces False Positives with High-Precision AI Detection: A trained model extracts feature images from subject video data and precisely determines the match with desired scenes. This could reduce false positives by up to 90% compared to conventional manual verification or simple sensor detection.

02

Cuts Operational Costs by One-Third Through Automated Scene Recording: Automatically detecting and recording specific scenes from vehicle footage eliminates the need for extensive manual video data review. This is estimated to reduce labor costs for operations by up to one-third.

03

Secures Strong IP in a Highly Competitive Field: This robust technology secured patent approval by overcoming rejection notices in a highly competitive area, citing over 10 prior art documents. It offers a clear advantage over existing technologies, providing a strong market differentiator.

Market Opportunity
Dashcams and ADAS Systems
$300M–$400M globally (AI est.)
Increasing awareness of accident reduction and the proliferation of Advanced Driver-Assistance Systems (ADAS) are driving demand for more sophisticated event recording and analysis capabilities.
Automotive electronics manufacturers ADAS software developers Dashcam hardware providers
Fleet Management Solutions
$150M–$250M globally (AI est.)
The logistics industry faces urgent challenges in ensuring safe driving practices, improving operational efficiency, and reducing costs, necessitating real-time AI-powered vehicle monitoring.
Logistics and transportation companies Telematics service providers Fleet software developers
Automotive Insurance Telematics
$400M–$500M globally (AI est.)
The growing adoption of telematics insurance increases demand for technologies that accurately capture driving behavior and accident circumstances, directly impacting premium optimization and fair claims assessment.
Automotive insurance providers Telematics data analytics firms Risk assessment software developers
IP Defensibility — Why Competitors Can't Replicate This
What This Patent Covers

This patent protects an AI-driven image detection system that uses a trained model to extract feature images from video data and determine a match against desired scenes. The claims were carefully refined and strengthened through the examination process, demonstrating clear differentiation from over 10 cited prior art documents and establishing a robust, low-invalidation-risk right.

Competitive White Space

This patent primarily covers AI-driven scene detection and event recording from vehicle video. White space exists in developing predictive analytics for driver behavior or vehicle maintenance, integrating with autonomous driving control systems, or applying the core AI model to non-vehicular moving object detection in diverse industrial settings.

Economic Impact
~$65K/year estimated economic impact per facility (est.)
estimated ROI · USD · AI analysis
ROI Calculation Logic

Assuming a company spends 200 hours/month on dashcam event scene detection, this technology could reduce that time by 80%. At an hourly labor cost of $33 (AI est.), this equates to an annual labor cost reduction of ~$65K (AI est.) per facility. High-precision detection also minimizes oversight risks, potentially generating an overall economic impact exceeding ~$65K (AI est.) annually.

Speed to Market
6× faster than in-house development
This technology has a strong technical foundation, with the trained model, feature image extraction algorithm, and event recording logic based on threshold determination all detailed in the patent specification. This significantly reduces development effort for licensees, allowing them to focus on integration into existing systems and potentially shortening time-to-market by approximately 2.5 years. Accelerated validation is expected.
Competitive Positioning

X: Detection Accuracy and Reliability
Y: Operational Efficiency and Cost-Effectiveness

Business Models & Applications
🤝 Technology Licensing
License the trained model and detection algorithms to dashcam manufacturers and automotive component suppliers, facilitating integration into their products.
☁️ SaaS Data Analytics Service
Offer a SaaS platform for fleet operators to analyze video data collected from vehicles in the cloud, providing automated reports on dangerous driving and specific events.
⚙️ AI Module Sales
Sell AI modules that can be easily integrated into existing vehicle systems and IoT devices, accelerating adoption across various industries and supporting new value creation.
Adjacent Application Opportunities
🚧 Construction and Civil Engineering
Heavy Equipment Proximity Detection System
AI could automatically detect workers, obstacles, or intrusions into hazardous zones from camera footage mounted on heavy construction equipment. This has the potential to enhance site safety and efficiency, reducing accidents by an estimated 20%.
🏭 Factory and Plant Operations
Production Line Anomaly and Quality Control
AI could analyze product images on manufacturing lines in real-time, automatically detecting defects, scratches, or foreign object contamination. This is expected to prevent defective products from reaching the market, improving overall production efficiency and quality by 15-20%.
🚨 Security and Surveillance
Intrusion and Abnormal Behavior Detection
AI could automatically detect intruders or abnormal behaviors (e.g., falls, altercations) from facility and wide-area surveillance camera footage, triggering alerts for security personnel and automated recording. This could reduce response times by 30% and enhance overall security levels.
Integration Roadmap — Estimated 12-Month Deployment
Phase 1: Technical Validation and Requirements Definition
Duration: 3 months
Verify compatibility with the licensee's existing camera systems and define detailed requirements for specific scenes or events to be detected. Develop an initial adjustment plan for the learning model and a data collection strategy.
Phase 2: Model Optimization and Prototype Development
Duration: 6 months
Optimize the learning model using collected data to meet the licensee's specific requirements and develop a prototype system. Conduct initial real-world testing to validate detection accuracy.
Phase 3: Implementation and Production Deployment
Duration: 3 months
Integrate the optimized model and system into the licensee's existing infrastructure and commence production operations. Based on post-deployment feedback, continuously improve the model and enhance performance.
Technical Feasibility
This technology is provided as a software-based trained model and detection algorithm, facilitating easy integration into existing vehicle-mounted cameras and dashcam systems. The image input, feature extraction, match determination, and event recording steps described in the patent specification can be implemented using general-purpose image processing libraries and AI frameworks, requiring no major hardware modifications. Deployment will primarily involve software updates or API integration, suggesting a relatively low technical barrier.
Success Scenario
Upon adopting this technology, fleet operators could automatically detect dangerous driving or accident precursors from vehicles and notify safety managers in real-time. This is estimated to reduce hundreds of annual manual video review hours by over 80%, leading to significant labor cost savings. Furthermore, high-precision event data could be utilized for optimizing insurance premiums and improving driver training programs, potentially reducing accident rates by 10%.
Patent Record
APPLICATION NO.
特願2024-023476
REGISTRATION NO.
7659929
FILING DATE
2024/02/20
GRANT DATE
2025/04/02
EXPIRATION DATE
2044/02/20
PATENT HOLDER
株式会社ユピテル
Examination History
2024年03月19日
出願審査請求書
2024年03月19日
手続補正書(自発・内容)
2024年12月03日
拒絶理由通知書
2025年02月03日
意見書
2025年02月03日
手続補正書(自発・内容)
2025年02月25日
特許査定