Time Source Ranking Using GPS Histograms for Sensor Timestamps
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous driving vehicles face challenges in accurately synchronizing time sources from various sensors and real-time clocks, leading to potential inaccuracies in motion planning and control due to the use of high precision crystal oscillators, which can be costly and not universally available, and confusion in time generation across multiple clock sources.
Innovation Solution
A method and circuit for ranking time sources in an autonomous driving vehicle by generating difference histograms between a GPS sensor and other sensors and real-time clocks, selecting the time source with the least difference, and generating timestamps based on the selected source to synchronize sensor data, ensuring accurate timekeeping and synchronization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If high precision crystal oscillators are used to generate time, then time accuracy is improved, but cost increases and availability decreases
Solution Approach 1:
The patent replaces expensive high-precision crystal oscillators with cheaper, more readily available clock sources. By using multiple lower-cost clock sources and selecting the best one through histogram analysis, the system achieves comparable time accuracy without the high cost and limited availability of precision oscillators.
Solution Approach 2:
The patent introduces a histogram-based selection mechanism as an intermediary between multiple clock sources and the final time output. This intermediary analyzes time differences from various sources and selects the most accurate one, enabling the system to achieve high precision without relying on expensive individual components.
2Adaptability or versatility
If multiple clock sources from sensors and devices are used, then time source availability is improved, but time generation becomes confusing and imprecise
Solution Approach 1:
The patent segments the time source selection process into distinct components: collecting times from multiple sources, generating separate difference histograms for each source, and independently analyzing each histogram. This segmentation allows systematic comparison and selection of the best time source while maintaining precision.
Solution Approach 2:
The patent implements feedback through histogram analysis, where time differences from multiple sources are continuously measured, analyzed, and used to select the optimal clock source. This feedback mechanism resolves the confusion of multiple sources by providing objective criteria for selection based on actual performance data.
Data Source
Figure 1
Figure 2
Figure 3A
AI summary
In one embodiment, a system receives a number of times from a number of time sources including sensors and real-time clocks (RTCs), wherein the sensors are in communication with the ADV and the sensors include at least a GPS sensor, and where the RTCs include at least a central processing unit real-time clock (CPU-RTC). The system generating a difference histogram based on a time for each of the time sources for a difference between a time of the GPS sensor and a time for each of the other sensors and RTCs. The system ranks the sensors and RTCs based on the difference histogram. The system selects a time source from one of the sensors or RTCs with a least difference in time with respect to the GPS sensor. The system generates a timestamp based on the selected time source to timestamp sensor data for a sensor unit of the ADV.