Context-Aware Quantitative Risk Assessment Machine Learning Model for Drivers Distraction
URL: http://arxiv.org/abs/2402.13421v1
archive: archived pipeline: cataloged verified
Abstract
Risk mitigation techniques are critical to avoiding accidents associated with driving behaviour. We provide a novel Multi-Class Driver Distraction Risk Assessment (MDDRA) model that considers the vehicle, driver, and environmental data during a journey. MDDRA categorises the driver on a risk matrix as safe, careless, or dangerous. It offers flexibility in adjusting the parameters and weights to consider each event on a specific severity level. We collect real-world data using the Field Operation Test (TeleFOT), covering drivers using the same routes in the East Midlands, United Kingdom (UK). The results show that reducing road accidents caused by driver distraction is possible. We also study the correlation between distraction (driver, vehicle, and environment) and the classification severity based on a continuous distraction severity score. Furthermore, we apply machine learning techniques to classify and predict driver distraction according to severity levels to aid the transition of control from the driver to the vehicle (vehicle takeover) when a situation is deemed risky. The Ensemble Bagged Trees algorithm performed best, with an accuracy of 96.2%.
Summary
Multi-Class Driver Distraction Risk Assessment (MDDRA) model that classifies a driver on a three-level matrix (safe / careless / dangerous) using vehicle, driver, and environment features drawn from the TeleFOT naturalistic dataset (East Midlands, UK). Driver features include PERCLOS, head turn/tilt, blink frequency, and yawning; vehicle and environment features include speed, weather, and road type. The model is intended to support takeover decisions when a situation is deemed risky.
Key finding
An Ensemble Bagged Trees classifier achieved 96.2% accuracy on the three-level distraction-severity classification, with PERCLOS and head pose contributing most to the dangerous-class boundary.
Methodology
Naturalistic driving data from the TeleFOT Field Operation Test on shared East Midlands routes. Features extracted across driver (PERCLOS, head pose, blink, yawn), vehicle (speed, maneuver), and environment (weather, road) channels. Multiple machine-learning classifiers compared; Ensemble Bagged Trees selected as best performer for the safe/careless/dangerous risk matrix.
Sample size: TeleFOT naturalistic dataset; specific driver count not extracted
Quality score: 5 / 5