A framework for reproducible design, implementation, and evaluation of driving-simulator experiments in SimCCAD for mixed traffic research
Year: 2026
Type: article
Abstract: Reproducibility remains a persistent challenge in transportation research, particularly in mixed-traffic studies involving connected and automated vehicles (CAVs). Driving simulators support controlled investigation of human–CAV interactions, but their experimental setups are often difficult to reproduce. Scenario geometry, traffic actors, controller logic, data logging, and pretesting procedures are frequently implemented through manual, bespoke, or poorly documented workflows. These limitations make experiment iteration time-consuming and error-prone, and they limit systematic assessment of whether a scenario is technically feasible under real-time human-in-the-loop operation. This paper proposes a reproducible design-to-implementation framework for mixed CAV driving-simulator experiments. The framework treats reproducibility as part of experiment design by linking parameterised experiment specification, scripted scenario generation, controller assignment, structured logging, pretesting, and benchmark-based feasibility checking. The framework is integrated and demonstrated within SimCCAD, a simulation platform for Conventional, Connected, and Automated Driving, through a scenario designed to investigate driver interactions with CAVs. The implementation shows how road geometry, traffic actors, controller logic, world files, and benchmarking outputs can be generated within an auditable project structure. The framework enables candidate scenario configurations to be generated, executed, and evaluated systematically under varying surrounding-CAV settings. This creates a reproducible basis for identifying feasible experiment configurations before participant deployment. The resulting workflow connects experiment design, executable simulator implementation, and benchmark-based feasibility evaluation within SimCCAD/Webots. Although the implementation is platform-specific, its underlying principles may also support mixed CAV studies on other platforms that provide equivalent mechanisms for parameterised design, scripted execution, documented refinement, and feasibility evaluation.
Language: English
Subfield: Control and Systems Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: gold
Estimated APC paid: $1,990
OpenAlex ID: W7225824868