Vehicle Action Determiner Using EEG Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face limitations in feedback quantity and quality, hindering the improvement of their machine learning driving capabilities, as they rely on limited user input for performance enhancement.
Innovation Solution
A method and system that detect user focus and brain activity to identify negative reactions, allowing autonomous vehicles to determine and correct undesired actions without explicit user feedback, utilizing electroencephalography (EEG) and control parameters to trigger corrective actions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If autonomous vehicles rely on user input for performance enhancement, then machine learning driving capabilities can be improved, but the quantity and quality of feedback remain limited
Solution Approach 1:
The system implements a feedback mechanism where users provide feedback about undesired vehicle actions through a user interface. This feedback is received by a server, processed to identify patterns of undesired actions, and used to update machine learning models that control autonomous vehicles, thereby continuously improving their driving capabilities through accumulated feedback data
Solution Approach 2:
A server acts as an intermediary between users and autonomous vehicles. The server collects feedback from multiple users, processes this data to identify undesired actions, and transmits updated machine learning models to vehicles. This intermediary structure enables centralized processing and sharing of feedback across the vehicle fleet, overcoming individual vehicle limitations
2Productivity
If autonomous vehicles collect and process more feedback data, then machine learning performance improves, but the system complexity increases
Solution Approach 1:
The system enables users to self-report undesired actions through a simple user interface without requiring complex automated detection systems in each vehicle. Users autonomously provide feedback about observed undesired actions, which the server then processes centrally, reducing the complexity burden on individual vehicles while maintaining high feedback quality
Solution Approach 2:
The server creates and distributes copies of updated machine learning models to multiple autonomous vehicles. Instead of each vehicle independently processing all feedback data, the server consolidates processing and replicates the results across the fleet, significantly reducing individual vehicle complexity while maintaining collective learning capability
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enables continuous improvement of autonomous vehicle performance by providing valuable feedback without user input, enhancing the learning capabilities of machine learning models and reducing the risk of undesirable actions.
Implementation Method 1
acquiring brain activity data of the first user; the brain activity data may be based on electroencephalography, EEG
Data Source
AI summary
It is presented a method performed in a vehicle action determiner for determining an undesired action of a first vehicle and a resulting action. The method comprises: detecting that a first user is focusing on the first vehicle; acquiring brain activity data of the first user; determining a negative reaction of the first user based on the brain activity data; determining when the first vehicle is performing an undesired action based on the first user focusing on the first vehicle and the negative reaction of the first user; determining the resulting action taken by an autonomous vehicle, based on the first vehicle performing the undesired action; and triggering the autonomous vehicle to perform the resulting action.