Driver Interruptibility Assessment Using Sensor Fusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for managing driver attention during driving tasks are inadequate, particularly when dealing with peripheral tasks, as they either disrupt primary tasks or fail to account for dynamic driving conditions, leading to safety concerns and inefficient user experience.
Innovation Solution
A system utilizing sensors to continuously monitor driver states and driving conditions, building a machine learning classifier to determine driver interruptibility in real-time, allowing for informed mediation of interruptions and optimal timing for peripheral interactions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If drivers perform peripheral tasks while driving, then user experience and information delivery improve, but driving safety and primary task performance deteriorate
Solution Approach 1:
The system performs preliminary assessment of driver interruptibility using sensor data (eye tracking, physiological sensors, steering wheel sensors) before allowing peripheral tasks to proceed. This advance evaluation enables the system to predict whether the driver can safely handle interruptions, and pre-regulate information flow accordingly, preventing safety-critical interruptions before they occur.
Solution Approach 2:
The patent introduces an intelligent intermediary system that mediates between external information sources and the driver. This intermediary assesses driver state in real-time and selectively filters or delays peripheral information based on safety considerations, acting as a buffer that allows both information delivery and safety maintenance.
2Loss of information
If self-report methods are used to sample driver experience, then user experience data is collected, but driver attention and cognitive capabilities for primary task deteriorate
Solution Approach 1:
The patent replaces the mechanical/self-report method of experience sampling with an automated sensor-based measurement system. Eye tracking sensors, physiological sensors, and steering wheel sensors automatically capture driver state data without requiring driver participation or attention, substituting human self-reporting with instrumental measurement.
Solution Approach 2:
The system performs self-measurement of driver state through embedded sensors that continuously monitor physiological indicators, eye movements, and steering behavior. The driver does not need to actively report their state; the system autonomously collects and processes this data to assess interruptibility.
3Measurement precision
If sensor data is continuously collected to assess driver state, then measurement precision of driver interruptibility improves, but device complexity and data processing requirements increase
Solution Approach 1:
The patent combines multiple sensor data streams (eye tracking, physiological sensors, steering wheel sensors, vehicle motion sensors) into a unified assessment model. By merging these diverse data sources and processing them through an integrated machine learning classifier, the system achieves high measurement precision while managing complexity through consolidation rather than separate processing systems.
Solution Approach 2:
The sensor system is designed with multi-functionality, where a single integrated platform performs multiple assessment functions simultaneously. The same sensor array that monitors driver attention also detects physiological stress, steering behavior, and vehicle context, eliminating the need for separate specialized systems for each measurement function.
Data Source
AI summary
The system described herein provides a sensor-based assessment of attention interruptibility. In a driving scenario, this system solves a persistent safety issue with regard to dangerous interruptions as they increase driver workload and reduce performance on the primary driving task. Being able to identify when a driver is interruptible is critical for building systems that can mediate these interruptions. The present invention utilizes sensor data collected from wearable devices and on-board diagnostics which can be used as input to build a machine learning classifier that can determine driver interruptibility in real-time.