Crowd-Sensed Traffic Light Mapping for Autonomous Vehicles
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous and semi-autonomous vehicles face computational inefficiencies when identifying traffic lights at intersections, as existing methods are time-consuming and computationally expensive, requiring rapid and accurate localization to follow traffic rules.
Innovation Solution
A system utilizing crowd-sourced data from probe vehicles to create and update digital maps with observed nodes, allowing host vehicles to locate traffic lights more efficiently by using mapped nodes with confidence coefficients and utility values, and removing outdated nodes, thereby reducing computational load and time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the vehicle uses computational processes to locate the traffic light from image data, then the traffic light can be identified accurately, but the computational time and resources increase significantly
Solution Approach 1:
The system pre-processes image data from multiple probe vehicles to create a digital map with pre-identified traffic light locations and characteristics. When a host vehicle approaches an intersection, the system queries this pre-built digital map instead of performing real-time computational analysis, thereby obtaining accurate traffic light location information without incurring high computational costs or time delays at the moment of decision-making
Solution Approach 2:
The system creates a digital copy of the physical intersection environment by processing image data from probe vehicles into a digital map representation. This digital map contains copied information about traffic light positions, colors, and other intersection features, allowing host vehicles to query pre-computed results rather than performing original computational analysis, thus achieving both accuracy and efficiency
2Reliability
If the vehicle performs real-time computational analysis of image data, then the traffic light state can be determined, but the computational resources and processing power increase
Solution Approach 1:
The system extracts and isolates the computationally intensive task of traffic light detection and state determination from the host vehicle's real-time processing requirements. Instead, probe vehicles perform this extraction and upload results to build a digital map, allowing host vehicles to query pre-computed states with minimal energy consumption while maintaining high reliability through the use of crowd-sourced data from multiple sources
Solution Approach 2:
The digital map serves as an intermediary between probe vehicles that capture image data and host vehicles that need traffic light information. The map pre-processes and stores traffic light states from multiple probe vehicles, acting as a mediator that eliminates the need for host vehicles to perform energy-intensive real-time computational analysis while ensuring reliable state identification through aggregated data
3Measurement precision
If the system uses crowd-sourced data from multiple probe vehicles, then the digital map accuracy improves, but the data processing and integration complexity increases
Solution Approach 1:
The system merges image data and traffic light observations from multiple probe vehicles into a unified digital map representation. By combining data from multiple sources that observe the same intersection, the system improves digital map accuracy through data aggregation and cross-validation, while the remote server handles the merging complexity centrally rather than requiring complex integration at each vehicle
4Measurement precision
If the system maintains updated digital maps with frequent observations, then the traffic light location accuracy improves, but the computational load for map updates increases
Solution Approach 1:
The system continuously updates the digital map with new observations from probe vehicles as they traverse intersections, maintaining current and accurate traffic light location data. This continuous update process improves positional accuracy over time through repeated measurements, while the remote server handles update operations asynchronously, preventing bottlenecks and maintaining high map update efficiency without compromising productivity
Data Source
AI summary
A system and method for locating a traffic light at a host vehicle. A probe vehicle obtains image data on an intersection that includes the traffic light when the probe vehicle is at the intersection and identifies the traffic light in the intersection from the data. A remote processor creates an observed node in a digital map corresponding to the traffic light and updates a mapped position of a mapped node within the digital map based on an observed position of the observed node. The host vehicle uses the mapped node of the digital map to locate the traffic light when the host vehicle is at the intersection.