In-Vehicle Driver Identification Using Driving Behavior Profiles

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In-vehicle computing systems face challenges in accurately identifying the driver based on driver-dependent functions, often relying on transferable items like key fobs or mobile devices, which can lead to misidentification.

Innovation Solution

The system detects driver-dependent functions by inferring the driver's identity through current driving behavior compared to generated driver profiles, using a weighted combination of data from vehicle sensors, external devices, and cloud-based networks to determine a confidence score for accurate identification.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If transferable items like key fobs or mobile devices are used for driver identification, then the system can perform driver-dependent functions, but misidentification occurs when these items are transferred between drivers

Engineering Contradiction:
Improvedriver identification accuracyVSAvoididentification system complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent replaces mechanical/physical identification methods (key fobs, mobile devices) with a behavioral biometric system that uses vehicle sensor data to infer driver identity. The system analyzes driving behavior patterns such as steering inputs, acceleration, and braking characteristics to automatically identify the driver, eliminating the need for transferable physical tokens that can be misplaced or shared.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Measurement precision

If driver profiles are generated using multiple data sources with weighted combinations, then identification accuracy improves, but processing complexity increases

Engineering Contradiction:
Improvedriver identification precisionVSAvoiddata processing complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent changes the parameters used for identification from static physical tokens to dynamic behavioral parameters. Multiple vehicle sensors collect data on driving patterns, and the system applies weighted combinations to these parameters to generate confidence scores for driver identification. The weighting allows the system to prioritize more reliable indicators while maintaining computational feasibility.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The system performs self-identification by automatically analyzing its own sensor data without requiring external authentication devices. The vehicle's existing sensor network serves the dual purpose of vehicle operation monitoring and driver identification, eliminating the need for separate identification hardware and reducing overall system complexity.

Inventive Principle:
Principle #25Self-service

3Adaptability or versatility

If the system selectively enables driver-dependent functions based on inferred identity, then notifications are more relevant to the driver, but false identification may lead to incorrect function activation

Engineering Contradiction:
Improvenotification customizationVSAvoidfunction activation reliability
Core Design Contradiction:
Adaptability or versatilityVSReliability

Solution Approach 1:

The system implements feedback through confidence score thresholds and verification mechanisms. When driver behavior is inferred, the system calculates a confidence score based on how well the observed behavior matches stored driver profiles. If the confidence score exceeds a threshold, the identification is confirmed and driver-dependent functions are activated; otherwise, the system may request alternative verification or default to a safe state, preventing false activation.

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP2891589B1Automatic driver identification
Publication Date: 2024.09.25 HARMAN INT IND INC
  • EP2891589B1 patent drawingFigure 1
  • EP2891589B1 patent drawingFigure 2
  • EP2891589B1 patent drawingFigure 3

AI summary

Method for an in-vehicle computing system (204) including detecting a request to perform a driver-dependent function (216), and inferring an identity of the driver based on current driving data/behavior relative to one or more driver profiles (206), each of the one or more driver profiles (206) generated using driver-specific past driving data/behavior. The method may further include selectively enabling performance of the driver-dependent function (216) of the in-vehicle computing system (204) based on the inferred driver identity.