Car-following Characteristics of Adaptive Cruise Control from Empirical Data

Goodall, Noah J.; Lan, Chien-Lun · 2020 · Crossref

DOI: 10.31224/osf.io/z8ktu

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 addresses the lack of empirical data for modeling the car-following behavior of automated vehicles, specifically focusing on Adaptive Cruise Control (ACC). As governments update capacity models to account for computer-driven vehicles, existing methods remain reliant on human driving behaviors, which differ significantly in perception, reaction time, and control precision. The research aims to establish accurate car-following models for a production vehicle with ACC to improve traffic microsimulation accuracy. The authors conducted field tests using a 2017 Audi Q7 equipped with ACC, driven on public roads in live traffic. Data collection involved a smartphone GPS sensor for speed and acceleration, a laser scanner for measuring gaps to the lead vehicle, and synchronized dash-mounted video cameras to monitor ACC status and traffic conditions. The study measured four key characteristics: standstill distance, startup time, unimpeded acceleration profiles, and deceleration profiles. Two car-following models were calibrated to this empirical data: the Intelligent Driver Model (IDM), specifically an enhanced version designed for ACC systems, and the Wiedemann 99 model, which is the default in the widely used VISSIM simulation software. The empirical results revealed specific behavioral traits of the test vehicle. The standstill distance was measured at approximately 3.5 meters, and the average startup time—the delay between the lead vehicle’s movement and the ACC vehicle’s acceleration—was 1.59 seconds. The unimpeded acceleration profile showed that the vehicle exceeded the ISO 15622 standard’s maximum comfortable acceleration of 2.0 m/s² at lower speeds. Regarding deceleration, 98% of observed decelerations were less severe than -2.0 m/s². When comparing the empirical data to the models, the IDM and its enhanced ACC variant predicted overly cautious braking, decelerating severely when time gaps were under 3 seconds. In contrast, the actual ACC vehicle allowed headways to shrink to 1.3 seconds before braking. However, VISSIM simulations using the newly calibrated Wiedemann 99 parameters matched the empirical unimpeded acceleration tests closely. The significance of this work lies in providing the first set of Wiedemann 99 parameters for a specific production vehicle with ACC based on disclosed empirical data from public roads. By calibrating the widely used VISSIM model to real-world ACC behavior, the study offers transportation agencies and researchers a practical, low-cost method to integrate automated vehicle characteristics into existing traffic simulations. This facilitates more accurate estimates of how increasing ACC penetration will impact road capacity and traffic flow, addressing a critical gap in current transportation planning literature.

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-nvidia on 2026-08-10; verification: verified.

Topics

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

• following distance

Information type

What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).

• Theoretical Contribution: computational model