Motion Sickness Prediction Using Biosignal and Vehicle Data Fusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies struggle to accurately predict and control motion sickness in passengers due to limitations in consistently measuring biosignals, leading to inconsistent effectiveness of anti-motion sickness solutions and discomfort.
Innovation Solution
A system utilizing a deep learning model to process biosignals and behavior signals from wearable sensors and vehicle sensors, segmenting and labeling these signals to enhance prediction accuracy, and controlling vehicle systems like displays, air conditioning, and diffusers to mitigate motion sickness.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional biosignal measurement methods are used to predict motion sickness, then prediction can be performed, but measurement consistency is poor leading to inaccurate detection
Solution Approach 1:
The patent combines multiple biosignal types (EEG, ECG, EMG, EDA, temperature, respiration) with vehicle behavior signals (acceleration, steering angle, GPS) into a unified prediction system. This multi-source data fusion approach compensates for the inconsistency of individual biosignal measurements by cross-validating across multiple physiological indicators and contextual vehicle data, thereby improving both measurement reliability and prediction accuracy.
Solution Approach 2:
The deep learning model acts as an intermediary that processes and integrates raw biosignal data with vehicle behavior data. The model transforms inconsistent raw measurements into reliable motion sickness predictions by learning temporal patterns and relationships across multiple signal sources, effectively mediating between noisy sensor inputs and accurate prediction outputs.
2Measurement precision
If deep learning models with multiple sensors are used, then prediction accuracy improves, but system complexity increases
Solution Approach 1:
The deep learning model serves multiple functions simultaneously: it processes various biosignal types, integrates vehicle behavior data, performs temporal pattern recognition, and generates motion sickness predictions. This multi-functionality consolidates what would otherwise require separate processing systems into a single unified model, improving accuracy while managing complexity through functional integration.
Solution Approach 2:
The system uses the vehicle's existing sensors (acceleration, steering angle, GPS) to provide contextual behavior signals that complement biosignal data. This self-service approach leverages already-available vehicle data without requiring additional external sensors, thereby improving prediction accuracy while minimizing the increase in system complexity.
3Object-affected harmful factors
If multiple vehicle control systems are adjusted, then passenger comfort improves, but control system complexity increases
Solution Approach 1:
The system dynamically adjusts multiple vehicle control parameters (steering angle, acceleration, air conditioning, lighting, music) based on real-time motion sickness predictions. Rather than using a fixed control strategy, the system adaptively modifies each control parameter according to the passenger's physiological state and vehicle context, improving comfort while managing complexity through dynamic rather than static control.
Solution Approach 2:
The system applies different control strategies to different vehicle systems based on their specific functions and the nature of motion sickness symptoms. For example, steering and acceleration adjustments address vestibular disturbances, while air conditioning and lighting modifications address sensory overload. This localized approach to control allows targeted intervention in each subsystem rather than uniform control across all systems.
Data Source
AI summary
A method including measuring a biosignal of a passenger in a moving device through a biosensor, acquiring a behavior signal of the moving device from a sensor of the moving device, inputting the measured biosignal and the acquired behavior signal to a processor including a deep learning model, segmenting, by the processor, the input behavior signal into units of segments and labeling the input biosignal, extracting, by the processor, a feature value by fusing the segmented behavior signal and the labeled biosignal, and controlling, by the processor, the moving device by predicting a motion sickness state of the passenger based on the extracted feature value.