AI Co-Driver and AR Integration for Dynamic Emergency Navigation and Collision Avoidance

Charissis, Vassilis; Khan, Soheeb; AlTarteer, Samar; Lagoo, Ramesh · 2026 · Crossref

DOI: 10.1109/icce67443.2026.11449652

archive: archived pipeline: cataloged verified

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

Summary

This paper addresses the problem of cognitive overload and "infobesity" in modern vehicles, where excessive data from multiple displays hinders driver decision-making and safety. The authors propose an integrated system combining an AI Co-Driver with a full-windshield Augmented Reality (AR) Head-Up Display (HUD). The system is designed to filter and prioritize critical traffic and environmental data, projecting essential guidance directly into the driver’s line of sight alongside spatial audio cues. This approach aims to reduce cognitive strain, improve situational awareness, and facilitate a gradual transition toward autonomous vehicle acceptance by building driver trust through supportive, rather than intrusive, assistance. The study evaluated the prototype using a high-fidelity Virtual Reality (VR) driving simulator featuring a 1:1 scale Mercedes A-Class cockpit and a CAVE (Cave Automatic Virtual Environment) setup. The simulation replicated a 28-mile motorway triangle in central Scotland, known for high collision rates, allowing for dynamic adjustments to traffic and weather conditions. Twenty licensed participants aged 18–60 underwent a five-stage evaluation: pre-test demographics, familiarization, and two counterbalanced experimental conditions. In one condition, drivers used traditional Head-Down Displays (HDD); in the other, they utilized the AI/AR HUD system. Each session included two embedded accident scenarios to test collision avoidance. Metrics such as speed, lane positioning, and collision events were recorded at 0.03-second intervals, supplemented by post-test questionnaires assessing user experience and trust. Results demonstrated a significant reduction in collisions when the AI/AR system was active. Drivers experienced 21 collisions in the unsupported HDD condition compared to only 4 collisions with the AI/AR assistance. Analysis of lane changes and maneuver timing revealed that the assisted group executed smoother, more controlled responses, such as gradual braking and timely lane changes, whereas unassisted drivers often remained on collision courses. Participant feedback indicated that the system felt intuitive and calming, comparable to having a knowledgeable companion. However, a subset of participants collided in both conditions, suggesting limitations related to individual reaction times and hazard perception that the system could not fully mitigate. The findings confirm that integrating AI with immersive AR visualization effectively reduces cognitive load and improves safety outcomes in high-risk scenarios. The system enhances situational awareness by delivering prioritized, context-sensitive guidance without diverting visual attention from the road. The authors conclude that this technology offers a viable pathway for building driver trust and facilitating the gradual integration of autonomous functionalities. Future work is recommended to explore adaptive response thresholds and automated transitions to temporary autonomous control when human reaction times prove insufficient, thereby addressing the limitations observed in slower-reacting drivers.

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-prismaquant on 2026-06-26; verification: verified.

Topics

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

• hud ar windshield