Overview
This research proposes an AI-enabled framework designed for the discovery and activation of technical knowledge contained within patent archives, particularly focusing on patents that have expired or are lapsing. The framework aims to identify technology trends from these archives and translate patent disclosures into actionable business pathways. These pathways are defined as structured commercialization routes, which can include SaaS products, services, licensing packages, consulting playbooks, training offerings, data products, or internal process tools. The approach treats patent expiry as a business signal and an archival transition, rather than primarily a legal issue, while acknowledging the importance of legal status as a risk-screening input.
Research Context
Patent databases represent a significant repository of public technical knowledge. However, the identification, interpretation, and reuse of this knowledge, especially after patent rights expire or lapse, often present challenges. The proposed framework addresses this by leveraging artificial intelligence to streamline the process of unearthing and contextualizing dormant technical information for potential commercial application.
The framework integrates multiple data streams and analytical techniques to achieve its objectives. These include patent metadata, maintenance-fee records, legal-status indicators, semantic search capabilities, and patent-family analysis. Additionally, market signals and generative AI workflows are incorporated to enhance the translation of technical information into business pathways.
Approach
The system architecture combines several components:
- Patent Metadata: Utilizing descriptive information about patents.
- Maintenance-Fee Records: Data indicating the status of patent upkeep payments.
- Legal-Status Indicators: Information regarding the legal standing of patents.
- Semantic Search: Enabling concept-based querying within patent texts.
- Patent-Family Analysis: Examining related patents across different jurisdictions.
- Market Signals: Incorporating external data relevant to commercial viability.
- Generative AI Workflows: Using AI to generate structured outputs for commercialization.
A proof of concept was conducted using all 378 records from an official weekly CIPO ST.96 archive. This process involved identifying 20 candidate patents that were either expired, lapsed, or nearing expiry. The stability of a transparent scoring model was tested, and a locally hosted Qwen3.6 model was employed to populate structured review packets. The review packets contained information pertinent to evaluating potential business pathways.
Findings
The evaluation of the proof of concept demonstrated several key outcomes:
- Reproducible Ingestion: The system successfully ingested patent records in a repeatable manner.
- Stable Rankings: The scoring model maintained stable rankings even under weight perturbation, suggesting robustness in its prioritization mechanism.
- Schema-Conformant Model Output: The generative AI model produced outputs that conformed to the designed schema for structured review packets.
- Incomplete Legal-Status Coverage: The evaluation identified limitations related to the completeness of legal-status coverage within the processed data.
- Need for External Review: The findings indicated a necessity for register review and expert human review to complement the AI processes, particularly concerning legal uncertainty, data limitations, and commercialization risk.
The research suggests that AI can function as a discovery and translation layer for technical knowledge that is otherwise dormant. However, it also highlights the critical need for such systems to explicitly represent and account for legal uncertainty, data limitations, and inherent commercialization risks.
Why This Matters
This work offers a framework for systematically leveraging the vast public archive of technical knowledge contained in expired or lapsing patents, potentially transforming underutilized information into viable business opportunities. By treating patent expiry as a business signal, the framework provides a structured approach for identifying and translating technical disclosures into defined commercial pathways. The findings underscore the potential for AI in this domain, while simultaneously emphasizing the necessity of human oversight and transparency regarding data limitations and risks in AI-driven innovation processes.