Driver Drowsiness Detection Using Steering and Lane Deviation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing driver monitoring systems for road vehicles face inaccuracies in determining drowsiness and attention levels due to calibration based on statistical and non-punctual user data, failing to account for individual differences and boundary conditions, leading to potential false positives and negatives.

Innovation Solution

A method utilizing sensor devices to detect steering and line parameters, processing a steering index and line index to calculate a driver's tiredness/attention level, with a sensitivity function to consider external boundary conditions, and alerting the driver when the attention level exceeds a threshold.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If facial monitoring systems are used to detect driver drowsiness, then driver attention can be monitored, but false positives and negatives occur due to individual differences in drowsiness manifestations

Engineering Contradiction:

Improvedrowsiness detection accuracyVSAvoidfalse positive rate

Core Design Contradiction:

Measurement precisionVSReliability

Solution Approach 1:

The system changes from using fixed statistical thresholds to dynamically adapting parameters based on individual driver baseline behavior. It monitors steering angle, steering speed, and lane position parameters over time to establish personalized norms, then detects deviations from these norms to identify drowsiness, thereby accommodating individual differences in drowsiness manifestations

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The system performs preliminary characterization of the driver's normal driving behavior before attempting to detect drowsiness. It collects and analyzes baseline data on steering patterns and lane positioning during periods when the driver is known to be alert, then uses this baseline to detect subsequent deviations that indicate drowsiness, improving both accuracy and reliability

Inventive Principle:

Principle #10Preliminary action

2Ease of manufacture

If statistical and non-punctual user data are used for system calibration, then system setup is simplified, but detection accuracy decreases due to inability to account for individual differences

Engineering Contradiction:

Improvesystem calibration simplicityVSAvoiddrowsiness assessment accuracy

Core Design Contradiction:

Ease of manufactureVSMeasurement precision

Solution Approach 1:

The system performs self-calibration by automatically characterizing each driver's baseline behavior through automated monitoring of steering parameters and lane position. The system collects data during initial driving periods, establishes personalized norms without manual intervention, and continuously adapts to individual driving patterns, eliminating the need for complex manual calibration while maintaining high detection accuracy

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

Inventive Principle:

Principle #10Preliminary action

3Device complexity

If boundary conditions are not considered in monitoring, then system complexity is reduced, but false alarms increase due to lack of context awareness

Engineering Contradiction:

Improvemonitoring system complexityVSAvoidalarm accuracy

Core Design Contradiction:

Device complexityVSReliability

Solution Approach 1:

The system dynamically adjusts monitoring sensitivity based on contextual parameters such as driving duration, time of day, and environmental conditions. It modifies the thresholds for triggering drowsiness alerts according to these boundary conditions, reducing false alarms during periods when drowsiness is more likely (e.g., nighttime, long drives) while maintaining appropriate sensitivity during other periods

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS20230356723A1Method for determining the drowsiness or attention level of the driver of a road vehicle, and road vehicle thereof

Publication Date: 2023.11.09 FERRARI SPA

AI summary

Method for determining the drowsiness/attention level of the driver of a road vehicle; the method comprises the steps of: detecting at least one steering parameter related to the kinematics of a driver-operated device for steering; where the steering parameter is related to the steering angle and/or the steering speed of the road vehicle; elaborating a steering index as a function of the at least one steering parameter detected; detecting at least one line parameter related to the position of the road vehicle within a lane; processing a line index as a function of the at least one line parameter detected; computing a driver drowsiness/attention index as a function of a combination of the steering index and the line index.