DReyeVR: Democratizing Virtual Reality Driving Simulation for Behavioural & Interaction Research

Silvera, Gustavo; Biswas, Abhijat; Admoni, Henny · 2022 · IROS 2022 / arXiv:2201.01931

archive: archived pipeline: cataloged verified

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

Summary

This paper introduces DReyeVR, an open-source virtual reality (VR) driving simulator platform designed to facilitate behavioral and interaction research involving human drivers. The authors address the high cost and inaccessibility of high-fidelity driving simulators, which typically require expensive 360-degree projection systems (CAVEs) costing approximately $300,000. While researchers often use cheaper flat-screen setups, these compromise immersion and field of view. DReyeVR leverages consumer VR hardware to provide a high-fidelity, immersive experience for under $5,000, enabling broader access to controlled driving experiments that are unsafe or impractical to conduct on real roads. Built upon the Unreal Engine and the CARLA autonomous vehicle simulator, DReyeVR adds specific features tailored for human subjects research. The platform integrates an HTC Vive Pro Eye headset for built-in eye and head pose tracking, capturing data at up to 120 Hz. To enhance ecological validity, the authors added spatially accurate audio cues, including engine sounds, pedestrian noise, and collision indicators, which are absent in the base CARLA simulator. The vehicle interface includes realistic elements such as side and rear-view mirrors and a heads-up display showing speed and gear status. Navigation is supported by dynamically placed in-world road signs rather than abstract waypoint coordinates. The system also features a custom sensor for recording and replaying driver inputs and gaze data, allowing for precise post-hoc analysis. Additionally, DReyeVR supports ROS compatibility and integrates with CARLA’s ScenarioRunner for defining complex traffic scenarios. The hardware setup consists of a VR headset, tracking stations, a force-feedback steering wheel, pedals, and an arcade seat, achieving a medium fidelity score of 9 out of 15 according to established simulator validation metrics. The software is modular, allowing for different hardware configurations, and supports Windows and Linux environments. The authors demonstrate the platform’s utility through an example study investigating Advanced Driver Assistance Systems (ADAS) interventions. In this scenario, drivers encounter a jaywalking pedestrian, and the system tests three conditions: auditory alerts, visual highlighting of the hazard, or automatic braking takeover. This setup allows researchers to measure objective metrics like reaction times and collisions while tracking eye gaze to assess situational awareness. The significance of DReyeVR lies in its ability to democratize high-quality driving simulation for academic and industrial researchers. By providing an affordable, open-source tool that captures critical behavioral signals like eye gaze and head pose, it enables rigorous studies on driver-vehicle interaction, trust in autonomy, and safety interventions. The platform bridges the gap between low-cost, low-fidelity simulators and expensive, high-fidelity systems, offering a robust environment for testing human factors in autonomous and assisted driving contexts.

Key finding

DReyeVR provides a low-cost, open-source VR-based driving simulation platform that achieves medium fidelity and supports comprehensive behavioral data collection, including eye tracking and head pose, for under 5000 USD.

Methodology

simulator

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. Discovered via qwen3.6_summarize on 2026-05-28 (3 acquisition events logged).

Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified_with_issues.

Topics

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

• simulator validity fidelity
• hud ar windshield
• simulator sickness
• visual occlusion
• situational awareness
• simulator training transfer

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).

• Methodological Resource: tool software, measurement protocol
• Theoretical Contribution: computational model