Frequency analysis reveals differences in autonomous and human driving
Quantifying Spectral Differences in Vehicle Between Production Autonomous and Human-Driven Vehicles Across Driving Scenarios
Robotics
Summary
Driving behaviors of self-driving cars and human drivers are different, but most studies look only at simple measures over time. The authors used a new approach that examines how driving signals change over different speeds and conditions like weather and traffic. They found that autonomous vehicles behave differently from human drivers depending on the situation, such as being more similar during car-following but more different when it’s rainy. This method helps better understand real-world driving differences.
What this means in practice
- •For automotive engineers: Assess autonomous vehicle control behavior across real-world conditions using frequency-based kinematic differences compared to human drivers.
- •For traffic management teams: Incorporate spectral driving behavior insights into traffic flow models to better represent autonomous vehicle impacts under varied scenarios.
Authors
Peiyi Fang, Xiangyu Li, Yonglin Weng, Ke Ma
Abstract
Differences in vehicle kinematic characteristics between production autonomous vehicles (PAVs) and human-driven vehicles (HVs) have been limitedly investigated by empirical studies. Most recent studies rely on simulation-based models, while some further investigate low-level adaptive cruise control (ACC) systems in controlled experiments. These methods commonly adapt some time-domain metrics to characterize PAV-HV differences across limited driving conditions. However, current PAVs equipped with high-level autonomous driving systems generate driving behaviors in a black box using data-driven models. These fundamentally different mechanisms for generating behaviors may produce distinct kinematic characteristics in traffic. More importantly, these time-domain metrics cannot reflect frequency-related traffic dynamics across different driving scenarios. Thus, this study adapted a real-world PAV dataset with four PAV platforms and developed a frequency-domain framework to quantify kinematic differences between PAVs and HVs across diverse driving scenarios, including varying driving states, lighting, weather, and vehicle densities. The framework transforms kinematic signals into the frequency domain and extracts spectral features, and then compares these features between PAVs and HVs based on kernel density estimation and Wasserstein distance. The results reveal clear scenario-dependent PAV-HV spectral differences. Specifically, speed-related differences were consistently smaller during car-following than cruising, while rainy conditions consistently enlarged acceleration-related differences compared with clear conditions. These findings highlight the necessity of multi-scenario evaluations and demonstrate the value of frequency-domain analysis for characterizing PAV-HV kinematic differences under real-world conditions.