Scholarlink

<p>Bird collisions with transmission lines (TLs) cause significant bird mortality worldwide, yet technological solutions targeting this infrastructure remain critically underdeveloped. This systematic review synthesizes 58 peer-reviewed articles and 86 patents on bird detection and collision mitigation technologies for TL infrastructure, following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines. Although designed to capture the state of the art for TLs, most of the retrieved literature and patents originated from the aviation and wind energy sectors, where economic loss and human safety drive investment rather than biodiversity conservation. This asymmetry constitutes the central finding of the review: among studies directly targeting TL contexts, 73.7% focus exclusively on detection, and only 8.1% of patents are explicitly oriented toward biodiversity conservation outcomes. Among studies addressing detection, deep learning architectures dominate, particularly convolutional neural networks (CNNs); You Only Look Once (YOLO) variants are the most frequently reported, though they are not patentable as open architectures. Patent analysis, restricted to TLs and substations, revealed a sigmoidal growth trajectory, with artificial intelligence and computer vision (AI/CV) patents approaching saturation (<i>K</i> = 33.53) by 2023 and passive mechanical deterrence patents by 2024 (<i>K</i> = 44.33), driven predominantly by Chinese filings (95.3% of patents). However, a persistent structural gap remains: TL-specific applications represent a minority of the evidence base, and AI models trained predominantly on aviation and wind-farm datasets exhibit domain-shift limitations that constrain transferability to TL contexts. Addressing this gap requires investment in TL-specific datasets, field validation, and policy frameworks that position bird conservation as a primary driver of innovation.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Monitoring birds interacting with power lines: a systematic review of detection technologies and the persistent gap in transmission line applications

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月25日

查看全文
  • Ícaro Menezes Pinto,
  • Luiggi Cavalcanti Pessôa,
  • Pedro Ricardo Nery Nunes Silva,
  • Gabriela Gama da Silva Santos,
  • Viviane Spencer Andrade,
  • Benito Rafael Santana de la Torre,
  • Pedro Emanuel Santos Machado,
  • Eduarda Silva Almeida,
  • Larissa Donida Biasotto,
  • Alessandra Schwertner Hoffmann,
  • Michele Ferreira Lima,
  • Ricardo Abranches Felix Cardoso Junior,
  • Beatriz Cabral Dias de Carvalho

摘要

Bird collisions with transmission lines (TLs) cause significant bird mortality worldwide, yet technological solutions targeting this infrastructure remain critically underdeveloped. This systematic review synthesizes 58 peer-reviewed articles and 86 patents on bird detection and collision mitigation technologies for TL infrastructure, following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines. Although designed to capture the state of the art for TLs, most of the retrieved literature and patents originated from the aviation and wind energy sectors, where economic loss and human safety drive investment rather than biodiversity conservation. This asymmetry constitutes the central finding of the review: among studies directly targeting TL contexts, 73.7% focus exclusively on detection, and only 8.1% of patents are explicitly oriented toward biodiversity conservation outcomes. Among studies addressing detection, deep learning architectures dominate, particularly convolutional neural networks (CNNs); You Only Look Once (YOLO) variants are the most frequently reported, though they are not patentable as open architectures. Patent analysis, restricted to TLs and substations, revealed a sigmoidal growth trajectory, with artificial intelligence and computer vision (AI/CV) patents approaching saturation (K = 33.53) by 2023 and passive mechanical deterrence patents by 2024 (K = 44.33), driven predominantly by Chinese filings (95.3% of patents). However, a persistent structural gap remains: TL-specific applications represent a minority of the evidence base, and AI models trained predominantly on aviation and wind-farm datasets exhibit domain-shift limitations that constrain transferability to TL contexts. Addressing this gap requires investment in TL-specific datasets, field validation, and policy frameworks that position bird conservation as a primary driver of innovation.