In-Vehicle Interface Using Eye Tracking for Personalization

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current in-vehicle computing systems lack seamless interoperability and personalized experiences across different environments, such as home, vehicle, and office, limiting user interaction and device communication, and failing to provide intuitive and adaptive interfaces.

Innovation Solution

The integration of eye tracking, facial recognition, and deep learning algorithms within an in-vehicle computing system to create a multi-modal human-machine interface that recognizes user intent and adapts the vehicle environment, providing personalized experiences and predictive maintenance.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If multiple sensing elements and processing systems are integrated to enable user recognition and personalized experiences, then user identification accuracy and personalization capability are improved, but device complexity increases

Engineering Contradiction:
Improvepersonalization capabilityVSAvoidsystem complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The system segments the personalization function into distinct modular components: eye tracking module, facial recognition module, voice recognition module, and contextual data processing module. Each module independently processes specific aspects of user identification and preference detection, reducing overall system complexity while maintaining comprehensive personalization capability.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The computing system integrates multiple sensing elements (cameras, microphones, sensors) that serve dual purposes: enhancing user identification accuracy while also detecting driver state for safety applications. This multi-functionality approach allows a single integrated system to address both personalization and safety needs without proportionally increasing complexity.

Inventive Principle:
Principle #6Universality (Multi-functionality)

2Speed

If real-time processing of multiple data streams from sensing elements is implemented, then responsiveness and user experience quality are improved, but computational energy consumption increases

Engineering Contradiction:
Improveprocessing speedVSAvoidcomputational energy consumption
Core Design Contradiction:
SpeedVSUse of energy by moving object

Solution Approach 1:

The system performs preliminary processing of sensor data streams using dedicated hardware accelerators and preprocessing filters before main computational analysis. Eye tracking data is processed through optical flow algorithms in real-time, while facial recognition images undergo preprocessing for feature extraction, reducing the computational burden on main processors and lowering energy consumption during intensive processing tasks.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system implements periodic sampling of sensor data streams at optimized frequencies rather than continuous full-rate processing. Eye tracking is sampled at frequencies sufficient for detection but below maximum camera capability, and facial recognition is triggered periodically or event-driven, reducing overall computational energy requirements while maintaining responsive performance.

Inventive Principle:
Principle #19Periodic action

3Measurement precision

If comprehensive user data collection and analysis are performed to predict user needs, then predictive capability and personalization accuracy are improved, but data privacy concerns and system complexity increase

Engineering Contradiction:
Improveuser intent recognition accuracyVSAvoiddata processing complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system introduces a contextualization engine as an intermediary layer between raw sensor data collection and user profile generation. This engine processes and filters data streams, extracting relevant contextual information while discarding redundant or sensitive data before it reaches the user profile database, thereby reducing data processing complexity while maintaining accurate user intent recognition.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The system implements on-device processing and analysis of user data, with the vehicle's computing system independently generating user profiles and detecting user intent without requiring constant cloud connectivity. This self-service approach processes data locally, reducing the complexity of data transmission and external processing infrastructure while maintaining high measurement precision for personalization.

Inventive Principle:
Principle #25Self-service

Data Source

PatentEP3482344B1Portable personalization
Publication Date: 2022.08.31 HARMAN INT IND INC
  • EP3482344B1 patent drawingFigure 1A
  • EP3482344B1 patent drawingFigure 1B
  • EP3482344B1 patent drawingFigure 2

AI summary

Embodiments are disclosed for an example in-vehicle computing system for a vehicle. In some embodiments, a personalized interactive experience within the vehicle is provided responsive to identification of a user in the vehicle. The user may be identified based on biometric data detection, which may include eye tracking, pupil monitoring, and head tracking in some examples. Features of the in-vehicle computing system may be selectively provided based on the identified user and/or learned data relating to the user and/or conditions of the vehicle.