Technology Category
Cognitive Modeling & Cognitive Systems: Adaptive BehaviorMultiagent Systems: Adversarial AgentsHumans and AI: Human-Aware Planning and Behavior Prediction
Drowsiness-Aware Adaptive Autonomous Braking System based on Deep Reinforcement Learning for Enhanced Road Safety
🤖 AI Summary
This study addresses the limitation of existing driver assistance systems in dynamically adapting braking strategies to real-time physiological states such as driver fatigue, which impairs accurate safety distance estimation. To overcome this, the authors propose an adaptive automatic emergency braking system that integrates vehicle dynamics with electrocardiogram (ECG) signals. The approach first employs a recurrent neural network (RNN) to recognize fatigue states from ECG data, then models the driver’s condition as action latency and incorporates it into the state space of a reinforcement learning framework. Braking decisions are optimized using a Double-Dueling Deep Q-Network (DQN). This work represents the first integration of physiological awareness with deep reinforcement learning for emergency braking, achieving a 99.99% collision-avoidance success rate in high-fidelity CARLA simulations and significantly enhancing system safety and adaptability under both fatigued and non-fatigued driving conditions.
Technology Category
Application Category
📝 Abstract
Driver drowsiness significantly impairs the ability to accurately judge safe braking distances and is estimated to contribute to 10%-20% of road accidents in Europe. Traditional driver-assistance systems lack adaptability to real-time physiological states such as drowsiness. This paper proposes a deep reinforcement learning-based autonomous braking system that integrates vehicle dynamics with driver physiological data. Drowsiness is detected from ECG signals using a Recurrent Neural Network (RNN), selected through an extensive benchmark analysis of 2-minute windows with varying segmentation and overlap configurations. The inferred drowsiness state is incorporated into the observable state space of a Double-Dueling Deep Q-Network (DQN) agent, where driver impairment is modeled as an action delay. The system is implemented and evaluated in a high-fidelity CARLA simulation environment. Experimental results show that the proposed agent achieves a 99.99% success rate in avoiding collisions under both drowsy and non-drowsy conditions. These findings demonstrate the effectiveness of physiology-aware control strategies for enhancing adaptive and intelligent driving safety systems.
Problem
Research questions and friction points this paper is trying to address.
driver drowsiness
adaptive braking
road safety
physiological state
autonomous driving
Innovation
Methods, ideas, or system contributions that make the work stand out.
deep reinforcement learning
driver drowsiness detection
physiology-aware control
adaptive autonomous braking
ECG-based RNN
🔎 Similar Papers
No similar papers found.
