Akademik Veri Yönetim Sistemi
Arabian Journal for Science and Engineering, 2025 (SCI-Expanded, Scopus)
Unmanned aerial vehicles (UAVs) have rapidly proliferated across diverse real-world applications, ranging from emergency response and surveillance to logistics and healthcare delivery. A primary challenge inherent in these applications is the task of UAV path planning, which involves determining feasible and optimal flight trajectories that minimize path length and energy consumption while adhering to various constraints. Despite extensive research having been conducted in the field of UAV trajectory planning and optimization, there is still no comprehensive survey that thoroughly explores the intersection of algorithmic trajectory planning methodologies, energy efficiency considerations, and practical deployment constraints across various application domains. This survey addresses this gap by reviewing more than 130 recent papers and proposing a novel taxonomy that bridges algorithmic approaches and application-specific requirements. We present a structured taxonomy covering application domains, algorithmic approaches, and optimization objectives, along with tables comparing strengths, limitations, and solutions. This is complemented by a quantitative synthesis for easily measurable evidence of key findings. We categorize UAV path planning according to applications, techniques, and objective criteria. Path planning techniques are further categorized into classical methods, meta-heuristic approaches, machine learning-based strategies, and hybrid methodologies. This classification highlights the way each approach balances the objectives of energy and distance. We also classify real-world UAV applications into four domains: emergency response, security and surveillance, environmental monitoring, and delivery and logistics. We analyze how domain-specific constraints drive different path planning priorities. An extensive comparison of the evaluation metrics and benchmarks used in the literature is presented, highlighting the necessity for standardized performance evaluation. The algorithmic approaches are distributed roughly as approximately 30% of the surveyed approaches are classical, 29% are meta-heuristic, 18% are AI-based, and 23% are hybrid approaches. Only 18% of studies report real-world experiments; most results are simulation only, emphasizing the lack of real-world experiment validation. We identify key technical challenges such as real-time computation, adaptation to dynamic environments, and coordination among multiple UAVs. Furthermore, we identify practical limitations that include battery endurance, communication barriers, and regulatory constraints. We also note evaluation gaps, such as the absence of standardized datasets and metrics. Finally, we present prospective avenues for further research, such as the integration of advanced AI and edge computing for smarter path planning, the creation of standardized testbeds, and the exploration of novel multi-domain applications. Additionally, we pose open research questions designed to steer the academic community’s efforts. This survey aims to serve as a definitive reference for researchers and practitioners seeking to design energy-efficient and distance-optimal UAV trajectories in real-world scenarios.