ADAS False Warning Reduction via On-Off Vehicle Data Sharing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Advanced Driver Assistance Systems (ADAS) face issues with false classification and threat assessment due to environmental noise and lack of location and history-based strategies, leading to unreliable performance.
Innovation Solution
The integration of on-vehicle and off-vehicle databases to store and share GPS coordinates and false output warnings, allowing the ADAS to develop a history of specific locations and adjust confidence levels to prevent false warnings, thereby improving classification and threat assessment accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If a non-location and non-history based classification strategy is used, then the system operation is simple, but false classification occurs between valid targets and noise
Solution Approach 1:
The system pre-collects GPS location data and environmental noise characteristics before actual target detection. By storing location-history pairs and noise profiles in advance, the system prepares classification criteria beforehand, enabling more accurate real-time classification without complicating the detection algorithm itself.
Solution Approach 2:
The system implements feedback loops where classification results are continuously compared against stored location-history data and noise profiles. When discrepancies are detected, the system adjusts classification thresholds and re-evaluates targets, creating a self-correcting mechanism that improves accuracy over time.
2Ease of operation
If a non-location and non-history based threat assessment strategy is used, then the system operation is simple, but inaccurate threat assessment occurs
Solution Approach 1:
The system pre-establishes threat assessment criteria by collecting and storing location-specific data including historical threat patterns and environmental characteristics. This preliminary data gathering enables the system to apply location-aware threat assessment rules without real-time complexity.
Solution Approach 2:
The threat assessment system continuously monitors detected targets against stored location-history threat patterns. When a target matches known threat signatures at specific locations, the system adjusts threat levels and triggers appropriate responses, creating a feedback-driven threat assessment mechanism.
3Device complexity
If the ADAS uses only on-vehicle database, then the system complexity is low, but feature performance improvement is limited
Solution Approach 1:
The system merges on-vehicle databases with off-vehicle cloud databases to create a unified knowledge base. Local GPS and sensor data are combined with remotely stored location-history pairs and environmental noise profiles, enabling comprehensive location-based classification and threat assessment while distributing data storage across multiple systems.
4Reliability
If the ADAS continuously updates knowledge base using off-vehicle communications, then feature performance improves over vehicle lifespan, but data communication and processing complexity increases
Solution Approach 1:
The system performs preliminary data filtering and validation before uploading to off-vehicle databases. By pre-processing sensor data, GPS coordinates, and classification results into standardized formats, the system reduces communication overhead and simplifies remote data integration while enabling continuous knowledge base updates.
Data Source
AI summary
An Advance Driver Assistance System of a vehicle collects measurements for a driving event, including GPS coordinates for a specific location from the EH system, and logs the data into an on-vehicle database. The ADAS uses the data in the on-vehicle database to develop a history associated with the specific location. Information, including GPS coordinates and recorded false output warnings, may be shared between the on-vehicle database and an off-vehicle database, such as associated with and Electronic Horizon system, providing an opportunity to significantly improve the feature performance of the ADAS. Driver assist information may be compiled between the off-vehicle database and the on-vehicle database, thereby continuously updating the knowledge base of the ADAS and optimizing feature performance of the ADAS over the life of the vehicle.