Vehicle Light Pattern Learning for Autonomous Driving
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems for monitoring vehicles in driver-assisted, highly automated, or autonomous driving rely on pre-known light patterns and movements, which are inadequate due to the diversity of vehicle types and light patterns, leading to potential safety issues and inefficiencies.
Innovation Solution
A method and system that detect and monitor vehicle lighting patterns and movements, comparing them to a database to identify new patterns and movements, and store them for future reference, allowing for continuous learning and improved safety in driving processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If pre-known light patterns and movements are used for monitoring, then the system complexity is reduced, but the reliability and adaptability decrease due to diversity of vehicle types
Solution Approach 1:
The system automatically learns and stores new light patterns and vehicle movements autonomously without requiring manual programming or updates. The monitoring system serves itself by continuously acquiring data, identifying new patterns, and expanding its own database, thereby adapting to diverse vehicle types while maintaining manageable complexity
Solution Approach 2:
The system implements continuous feedback loops where detected light patterns and vehicle movements are compared against stored data, and new patterns are fed back into the database for future recognition. This feedback mechanism enables the system to improve its reliability over time by learning from actual vehicle behavior while keeping the operational complexity low through automated pattern recognition
2Ease of operation
If pre-known light patterns are used for monitoring, then the ease of operation is improved, but the adaptability deteriorates due to diversity of vehicle types
Solution Approach 1:
The system automatically adapts to new vehicle types and light patterns through self-learning mechanisms, eliminating the need for manual updates or reconfiguration. This maintains ease of operation while significantly improving adaptability to diverse vehicle types and their varying lighting behaviors
Solution Approach 2:
The system transitions from a static database of pre-known patterns to a dynamic, evolving database that continuously incorporates new light patterns and vehicle movements. This dynamic adaptation allows the system to maintain ease of operation while becoming versatile across different vehicle types and scenarios
3Measurement precision
If continuous monitoring of vehicle lighting is performed, then the measurement precision is improved, but the loss of time increases due to pattern analysis
Solution Approach 1:
The system performs preliminary actions by continuously monitoring and storing light patterns and vehicle movements in advance. This preparatory data collection and pattern storage enables rapid comparison and identification during critical moments, improving measurement precision without causing time loss during actual driving scenarios
Solution Approach 2:
The system maintains continuous monitoring and learning operations throughout vehicle operation, ensuring that pattern recognition and data collection occur simultaneously with normal driving. This continuous useful action eliminates idle time for pattern analysis while maintaining high measurement precision through ongoing data acquisition and comparison
Data Source
Figure 1
Figure 2
Figure 3
AI summary
It is described a method for learning a vehicle behaviour of a monitored vehicle (100). The method comprises detecting (10) at least a part of a vehicle illumination (110) of the monitored vehicle (100) and monitoring (20) the detected (10) vehicle illumination (110). And if a light-pattern (111) occurs in the detected (10) vehicle illumination (110), wherein the light-pattern (HI) corresponds to a frequency, intensity and/or colour dependant glowing of the vehicle illumination (110), the light-pattern (111) starting with a flashing up of at least a part of the detected (10) vehicle illumination (110) and ending after a certain time without glowing of the respective part of the detected (10) vehicle illumination (110), the method moreover comprises: monitoring (40) the light-pattern (111); monitoring (60) a vehicle movement (120) of the monitored vehicle (100) during the occurrence of the light-pattern (111); and comparing (50) the monitored (40) light-pattern (111) with at least a known light-pattern (211) from a light-pattern data entry (210) stored in an light-pattern database (200). And if the comparison (50) results into the monitored (40) light-pattern (111) being unknown, the method moreover comprises: storing (80) the light-pattern (111) and the vehicle movement (120) together as a new light-pattern data entry (220) into the light-pattern database ( 200 ).