AI System Automatically Tracks How Products Evolve, Unlocking Patent Intelligence
Researchers have developed an AI tool that automatically identifies engineering design evolution patterns hidden in patent documents, achieving 83% accuracy. The breakthrough enables companies and policymakers to systematically monitor competitor innovation trends and forecast technology development—turning millions of patents into actionable business intelligence.
Originaltitel: Navigating patent trends through the lens of system components completeness in trend of engineering system evolution
This study addresses the challenge of automating the Trend of Engineering System Evolution (TESE) process within patent classification an area underexplored in the context of design research. Specifically, we focus on the Trend of Increasing Completeness of System Components, characterised by four key elements: Operating system, Transmission, Energy source, and Control system (OTEC). We propose a novel classification framework that leverages Natural Language Processing (NLP) techniques TF-IDF and Naive Bayes, within the CRISP-DM methodology to categorise patents based on OTEC components. Sewing machine patents serve as a case study, achieving a promising 83% classification accuracy through cross-validation. This AI-driven approach provides a scalable model for automating the recognition of evolutionary design trends embedded in patent literature. Our findings establish a foundation for future research at the intersection of engineering design evolution, computational classification, and design knowledge management, contributing to more adaptive and intelligent design support systems.