Autonix: Real-Time Driver Drowsiness and Crash Detection System

Shaikh, Maizah; -, Yatin Anchan; Phudinawala, Hasan · 2025 · Crossref

DOI: 10.63363/aijfr.2025.v06i05.1559

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 introduces Autonix, an Android-based mobile application designed to enhance road safety by providing real-time driver drowsiness and crash detection. The research is motivated by the high incidence of accidents caused by drowsy driving and the limited accessibility of Advanced Driver Assistance Systems (ADAS), which are typically restricted to premium vehicles due to high costs and specialized hardware requirements. Existing mobile solutions often suffer from single-parameter detection limitations, poor real-world adaptability, and inadequate emergency response protocols. Autonix aims to democratize safety technology by leveraging standard smartphone sensors to create an affordable, non-intrusive, and comprehensive driver assistance system. The system utilizes a multi-modal approach combining computer vision and motion sensor data. For drowsiness detection, Autonix employs facial landmark tracking via the front-facing camera to monitor eye closure duration, blink frequency, and facial fatigue indicators. It incorporates adaptive thresholds to learn individual driver patterns, aiming to reduce false positives. For crash detection, the application integrates accelerometer and gyroscope data to analyze velocity changes and distinguish genuine collisions from normal driving disturbances like hard braking or potholes. The architecture includes a multi-level alert system that escalates from audio and visual warnings to automatic notifications to emergency contacts if the driver becomes unresponsive, featuring a countdown override to prevent false emergency alerts. Additionally, the app provides an analytics dashboard for real-time performance tracking and historical data analysis. Prototype testing with a diverse group of drivers across various vehicle types and conditions demonstrated the system's effectiveness. The multi-parameter facial analysis achieved high accuracy in identifying drowsiness episodes while significantly reducing false positives compared to single-parameter systems. The motion sensor integration reliably distinguished crash events from routine driving impacts. Users reported high acceptance rates, citing the convenience and non-intrusive nature of the smartphone-based approach. The multi-level alert mechanism successfully warned drowsy drivers, and the analytics features helped users recognize dangerous patterns and improve driving behavior. The study confirms that Autonix is accessible to a broader audience than traditional ADAS systems, operating effectively on standard Android devices without additional hardware. The significance of this work lies in its ability to bridge the gap in road safety accessibility. By eliminating the need for expensive retrofits, Autonix makes advanced safety features available to everyday vehicle users. The system promotes proactive accident prevention through real-time monitoring and rapid emergency response. Future enhancements proposed include offline functionality, AI-based prediction of drowsiness onset, integration with wearable physiological sensors, voice assistant support, and community safety networks for identifying high-risk road segments. This research contributes to the field by demonstrating that hybrid, sensor-based mobile applications can provide robust, scalable, and affordable solutions for critical driver safety challenges.

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.

• drowsiness detection algorithms
• dms validation
• telematics crash prediction

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