Safety Through Agility: Using Mixed Reality to Tune Shared Autonomy Systems

Loeb, Helen · 2024 · ROSA P / Carnegie Mellon University. Traffic21 Institute. Safety21 University Transportation Center (UTC)

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 research addresses the challenge of maintaining mobility and independence for aging populations while ensuring safety through shared autonomy systems. As the number of older adults increases, revoking driving privileges often diminishes their quality of life. The authors propose a hybrid driving system that mediates control between human drivers and autonomous vehicles, allowing humans to retain control up to their performance level while the system intervenes to prevent unsafe states. The project aims to develop an interactive imitation learning framework and a "Guardian Angel" safety system that monitors driver cognitive and physical states to personalize autonomy levels. The methodology involves developing a mixed reality (MR) driving simulator to test these shared autonomy concepts safely. The hardware setup includes a Logitech G29 steering wheel and pedal set paired with a Meta Quest 3 MR headset, integrated with the CARLA simulation environment via ROS2. The system employs a Driver Monitoring System (DMS) using vision-based tracking (eye gaze, head pose, blink patterns) and radio-frequency physiological indicators to detect distraction, fatigue, and cognitive decline. To evaluate the simulator’s efficacy for driver education, the researchers conducted a pilot study with 23 participants aged 16–30 at a driving school. Participants performed parallel parking tasks in the simulator and completed post-experiment surveys assessing immersion, intuitiveness, usefulness, fun, and helpfulness. The findings from the pilot study indicate that participants viewed the MR simulator as a valuable educational tool. Likert scale mean scores placed all metrics in the "somewhat agree" to "agree" range, with "helpful" receiving the highest agreement (mean 1.63 on a 5-point scale where 1 is strongly agree). Specifically, 86% of participants agreed or strongly agreed that the simulator would be helpful for driver education. Correlation analysis revealed that perceived usefulness was strongly correlated with immersivity (0.79), intuitiveness (0.78), and fun (0.76). The results suggest that improvements in immersion and intuitiveness directly enhance the simulator’s perceived efficacy. Additionally, the technical development demonstrated the feasibility of tracking visual cues and driving behaviors, such as lane discipline and braking patterns, to build anomaly detection models for identifying impairment events. The significance of this work lies in its potential to bridge the gap between human trust and machine safety in autonomous vehicles, particularly for vulnerable populations like the elderly and novice drivers. By validating MR simulators as effective tools for training and assessment, the research supports the development of shared autonomy systems that can prolong driving independence while mitigating risk. The authors conclude that such hybrid systems, combined with realistic simulation for testing and education, are critical for the successful transition to autonomous vehicle technologies, ensuring that benefits are accessible across all generations.

Key finding

Mixed reality driving simulators are perceived as helpful for driver education, with their perceived usefulness strongly correlated with levels of immersion and intuitiveness.

Methodology

simulator

Sample size: 23

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 bulk_ingest_rosap on 2026-05-23 (6 acquisition events logged).

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

Topics

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

• situational awareness
• automation
• automation surprise
• in vehicle coaching

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, validation psychometrics
• Theoretical Contribution: computational model