Market Context — Why This Technology, Why Now

The global push for decentralized computing and real-time data processing at the edge is intensifying, driven by the proliferation of IoT, AI, and 5G technologies. Industries demand smaller, more powerful, and energy-efficient devices capable of complex computations without relying heavily on cloud infrastructure. This patent offers a critical solution to these challenges, enabling advanced AI models to run on resource-constrained hardware, thereby accelerating innovation in smart manufacturing, autonomous systems, and connected health.

Key Competitive Advantages
01

Reduces memory capacity by up to 30% compared to conventional methods by commonizing function definition values per group, significantly contributing to the miniaturization and power saving of edge devices.

02

Enhances computation efficiency by up to 20% through commonized function information, reducing processing load on multiple function computation units, establishing superiority in fields requiring real-time processing.

03

Demonstrates high technical stability and uniqueness, with patentability recognized against 5 prior art documents, representing a robust right that has overcome examiner objections, enabling early market share acquisition.

Market Opportunity
IoT Devices & Edge AI
$13.5B–$33.5B globally (AI est.)
There is a strong demand for increased data processing, device miniaturization, and power efficiency. This technology's memory optimization and computation acceleration directly enhance device performance in these areas.
Edge AI chip manufacturers IoT device OEMs Embedded system developers Smart sensor manufacturers
Autonomous Driving & Robotics
$13.5B–$33.5B globally (AI est.)
Real-time, high-precision non-linear function computation is critical for safety and performance. This technology offers low-latency and high-efficiency processing, making its adoption highly beneficial.
Autonomous vehicle system integrators Robotics control system developers ADAS component suppliers Drone technology companies
Industrial Control Systems
$1.5B–$3.5B globally (AI est.)
Optimizing manufacturing lines and predictive maintenance requires complex data analysis and real-time control. This technology's stable, high-efficiency computation could contribute to significant productivity improvements.
Industrial automation solution providers Predictive maintenance software vendors Smart factory equipment manufacturers Process control system developers
IP Defensibility — Why Competitors Can't Replicate This
What This Patent Covers

This patent protects a computing device and method that efficiently manages function definition values by grouping multiple function computation units and storing common function information. The claims have been strengthened through amendments, overcoming examiner objections, indicating a robust and stable right.

Competitive White Space

While this patent covers efficient function computation, it does not explicitly claim specific non-linear function types or advanced hardware architectures for neural network acceleration. A licensee could build additional IP around novel function implementations or specialized hardware integrations for specific AI workloads.

Economic Impact
~$200K/year estimated memory and power consumption cost savings per facility (est.)
estimated ROI · USD · AI analysis
ROI Calculation Logic

Assuming this technology is implemented in 100,000 IoT edge devices, the annual memory cost reduction per device due to reduced storage capacity is estimated at ~$1.35 (AI est.), and operational cost improvement due to power consumption reduction is estimated at ~$0.65 (AI est.). This leads to an expected direct cost reduction of ~$2.00 (AI est.) per device, totaling ~$200K/year (100,000 devices × ~$2.00/device). Additionally, improved computation efficiency could extend product lifecycles and create opportunities for new features.

Speed to Market
3× faster than in-house development
This technology establishes the logic for function information management in computing devices, with basic algorithms detailed in the patent specification. This significantly shortens the conceptual design and fundamental research phases compared to developing similar technology from scratch. By assuming integration into existing computing devices and software architectures, the development period could be reduced by approximately 2.0 years, enabling faster market entry.
Competitive Positioning

X: Computation Efficiency
Y: Resource Optimization

Business Models & Applications
📝 Technology Licensing Model
License this technology to IoT device manufacturers and AI chip developers. Promote integration into products to generate continuous royalty income.
🤝 Joint Development & Customization Model
Jointly develop custom computing devices or software based on this technology for specific industries or clients. Offer as a high-value-added solution.
📦 Embedded IP Solution Sales
Develop and sell proprietary computing modules or edge AI platforms with this technology embedded. Differentiate by emphasizing high performance and resource efficiency.
Adjacent Application Opportunities
🤖 Robotics
Real-time Control for Autonomous Mobile Robots
Complex non-linear function computations are essential for environmental perception and path planning in autonomous mobile robots. Implementing this technology could enable higher precision control with limited resources, extending battery life and improving response speed. It would be particularly valuable when integrating and processing data from multiple sensors, potentially reducing processing latency by 20%.
🧬 Medical & Healthcare
Biosignal Analysis in Wearable Devices
For small wearable devices analyzing biosignals (heart rate, blood glucose, etc.) in real-time, this technology's memory-saving and high-efficiency computation would be highly effective. This could enable longer device operation and the implementation of more sophisticated anomaly detection algorithms, potentially reducing power consumption by 30% and improving user experience.
📡 Communication Infrastructure
Edge Processing for 5G/Beyond 5G Base Stations
In next-generation communication systems, edge computing at base stations is crucial. This technology could optimize computational resources and power consumption for real-time processing of vast data streams within base stations, potentially improving data throughput by 20%. This would enable efficient accommodation of more users and enhance the stability of low-latency service delivery.
Integration Roadmap — Estimated 23-Month Deployment
Phase 1: Technical Evaluation & PoC
Duration: 5 months
Based on the patent specification, evaluate the applicability of this technology to the licensee's existing systems and conduct a Proof of Concept (PoC) for specific use cases. Verify memory reduction effects and computation efficiency improvements.
Phase 2: Prototype Development & System Integration
Duration: 9 months
Based on PoC results, develop a prototype system incorporating this technology. Proceed with integration into existing computing devices and software frameworks, conducting functional tests and performance evaluations.
Phase 3: Production Deployment & Optimization
Duration: 9 months
Following prototype validation, proceed with deployment into the actual operating environment. Establish maximum economic benefits and technical advantages through long-term performance monitoring and continuous optimization using real-world data.
Technical Feasibility
This technology pertains to the functional block configuration and data management logic of computing devices. It can be integrated into existing computing device architectures via software or firmware updates for the common function information storage unit and grouping logic. Designed for high compatibility with existing systems, it avoids extensive hardware modifications. This suggests that adopting companies could implement this technology with relatively low initial investment.
Success Scenario
Upon implementation, this technology could enhance edge device computation processing speed by up to 20% while simultaneously reducing memory capacity requirements by 30%. This could enable the deployment of more complex AI models on smaller devices, leading to new service opportunities and significant power consumption reductions. Ultimately, this is estimated to improve product competitiveness and establish market leadership.
Patent Record
APPLICATION NO.
特願2021-041855
REGISTRATION NO.
7609419
FILING DATE
2021/03/15
GRANT DATE
2024/12/23
EXPIRATION DATE
2041/03/15
PATENT HOLDER
国立大学法人横浜国立大学
Examination History
2023年12月27日
出願審査請求書
2024年10月01日
拒絶理由通知書
2024年11月20日
意見書
2024年11月20日
手続補正書（自発・内容）
2024年12月03日
特許査定