Evaluating a Signalized Intersection Performance Using Unmanned Aerial Data

Ashqer, Mujahid I.; Ashqar, Huthaifa I.; Elhenawy, Mohammed; Almannaa, Mohammed; Aljamal, Mohammad A.; Rakha, Hesham; Bikdash, Marwan · 2022 · Crossref

DOI: 10.21203/rs.3.rs-1859329/v1

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study presents a novel methodology for computing Measures of Effectiveness (MOEs) at signalized intersections using vehicle trajectory data collected by Unmanned Aerial Vehicles (UAVs). The research addresses the limitations of traditional traffic detection methods, such as inductive loops and sparse GPS probe data, which often lack the spatial resolution or real-time accuracy needed to characterize complex traffic dynamics. By leveraging drone footage, the authors aim to demonstrate that UAVs can provide high-fidelity, microscopic data capable of estimating critical performance metrics—including queue lengths, spillbacks, delays, fuel consumption, and crash risks—in real time. The experimental design utilized a 14-minute video recorded by a drone hovering over a busy three-way signalized intersection in Athens, Greece, as part of the pNEUMA open data initiative. The study area included three approaches with varying lane configurations and a speed limit of 55 km/h. Researchers processed the video to extract second-by-second trajectories for over 750 vehicles, including cars, motorcycles, buses, and heavy trucks. The methodology involved dividing the intersection into lane-specific sub-polygons using geographic data, assigning vehicles to lanes based on instantaneous coordinates, and constructing time-space diagrams. A microscopic shockwave analysis was then applied to these trajectories to identify queue formation and recovery waves. The study defined a vehicle as stopped if its speed was ≤1.2 m/s and used this threshold to detect spillbacks occurring at the upstream edge of the monitored area. The results demonstrate the successful estimation of various MOEs. The authors calculated maximum queue lengths, vehicle stops, travel times, and delays for each lane. They also estimated crash rates using a time-dependent exposure model based on US national statistics and computed fuel consumption and CO2 emissions using the Virginia Tech Comprehensive Power-Based Fuel Consumption Model (VT-CPFM), which relies on instantaneous vehicle power. Furthermore, the study calibrated the Van Aerde car-following model using the drone data to construct fundamental diagrams, addressing limitations in traditional models like Greenshields and Pipes. The analysis revealed detailed traffic behaviors, such as frequent lane changes and multiple stops per vehicle during oversaturated conditions, which macroscopic models typically overlook. The significance of this work lies in its validation of UAVs as a viable, low-cost, and minimally intrusive tool for real-time traffic monitoring. The microscopic approach allows for the capture of transient vehicle behaviors and individual interactions, providing more accurate MOEs than aggregate methods. The findings suggest that drone-based data can support advanced traffic management applications, enabling dynamic signal control and better assessment of intersection performance under varying saturation levels. This study contributes to the field by providing a comprehensive framework for converting raw aerial video into actionable traffic engineering metrics.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• exposure measurement