Research on Environmental Perception Technology in Unmanned Delivery Vehicles

Abstract

This paper comprehensively reviews the mainstream environmental perception technology systems for autonomous delivery vehicles, systematically analyzing the working principles, technical architectures, performance parameters, application scenarios, and advantages and disadvantages of four primary sensing technologies: LiDAR, visual cameras, millimeter-wave radar, and ultrasonic sensors. It delves into core algorithms, fusion architectures, and implementation processes of multi-sensor fused perception. Drawing on real-world deployment cases from leading companies such as JD.com, Meituan, and Nuro, and supported by authoritative 2025 industry statistics and market analysis in autonomous delivery, the study systematically summarizes current technological bottlenecks, scenario adaptation challenges, and engineering implementation gaps in environmental perception for autonomous delivery vehicles. Targeted optimization strategies are proposed, along with predictions on future technological evolution and industry development trends. The research provides theoretical foundations and practical references for optimizing perception systems, facilitating engineering deployment, and advancing industry-standardization efforts in autonomous delivery vehicles, while also offering reliable support for upgrading intelligent logistics equipment and advancing applied engineering research.