以專利分析法研究自動駕駛車輛外部環境感知技術

Abstract

隨著人工智慧技術提升，人工智慧的應用已逐漸從學術理論走入大眾的日常生活，車用領域更是人工智慧未來發展的重要方向，特別是在面對中國新興品牌的崛起，歐美日各國傳統車廠都已逐漸感受到不得不發展下一代新型車款壓力的當下，自動駕駛車輛的研究便是未來車廠的發展重點。本研究利用專利分析方法，分析自動駕駛車輛四大外部環境感知技術— 雷達感知技術、光達感知技術、攝影機感知技術及超音波感知技術，找出專利技術發展趨勢以及專利優勢明顯的國家及廠商，使用羅吉斯成長模型（Logistic Growth Model）分析技術生命週期，評估技術發展階段，以專利數據探討自動駕駛車輛技術未來發展趨勢。研究結果發現，雷達感知技術、光達感知技術、攝影機感知技術及超音波感知技術，這些技術正在從硬體發展走向軟硬體整合的路上，根據歷年專利數與技術生命週期分析顯示，相關硬體技術已進入或接近成熟階段，證實當前自動駕駛車輛硬體技術發展已相對成熟，對於車廠本身或車輛零組件相關產業而言，未來發展方向將更著重於人工智慧或相關演算法軟體的實際應用，同時進一步優化生產，降低相關硬體的製造成本，才有望維持產業競爭優勢。With advancements in artificial intelligence (AI) technology, its applications have gradually transitioned from academic theory to the everyday lives of the public. The automotive sector is a crucial direction for the future development of AI. As traditional car manufacturers worldwide face increasing pressure to innovate next-generation vehicles in response to the rise of new Chinese brands, research on self-driving vehicles has become a key focus for the future of the automotive industry.This study utilizes patent analysis methods to examine fourmajor external environment perception technologies for self-driving vehicles. The findings indicate that radar, lidar, camera, and ultrasonic perception technologies are shifting from hardware development to the integration of software and hardware. Analysis of historical patent numbers and the technology life cycle suggests that the related hardware technologies have reached or are approaching saturation, with patent numbers expected to decline over the next five years. This confirms that the hardware technology for self-driving vehicles is relatively mature. For car manufacturers and the vehicle component industry, future development will increasingly focus on the practical application of artificial intelligence and related algorithm software, while also optimizing production to further reduce hardware manufacturing costs, in order to maintain a competitive edge in the industry.