ADS Monitoring Layer for Fleet Event Data and Bandwidth Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current Automated Driving System (ADS) solutions face challenges in efficiently collecting and processing data from a fleet of vehicles due to bandwidth and storage limitations, making it difficult to maintain performance over time and requiring reliance on limited test vehicle data.
Innovation Solution
A computer-implemented method and system that separates data into event message and event recording datasets, allowing for efficient transmission and storage of metadata and perception/data, respectively, enabling continuous data collection from production fleets without overwhelming bandwidth or storage resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If data from an entire fleet of vehicles is accessed and transmitted to a centralized server, then data quantity and diversity are improved, but bandwidth consumption and storage requirements increase significantly
Solution Approach 1:
The patent segments the data transmission process into two distinct streams: event message data (metadata) and event recording data (perception and ADS output data). This segmentation allows the system to transmit only essential metadata continuously while storing larger perception datasets locally, significantly reducing bandwidth consumption while maintaining the ability to retrieve comprehensive data when needed.
Solution Approach 2:
The patent extracts and transmits only the critical metadata components (event messages) to the centralized server, while leaving the bulk perception data stored locally in the vehicle's datastore. This extraction approach enables the server to receive actionable information without being overwhelmed by the volume of raw sensor data, effectively addressing the bandwidth limitation.
2Quantity of substance
If data from an entire fleet of vehicles is accessed and transmitted to a centralized server, then data quantity and diversity are improved, but post-processing complexity increases
Solution Approach 1:
The patent applies preliminary action by organizing and structuring data into standardized event message formats with metadata before transmission to the server. This pre-organization includes categorizing events by type (hazardous, informative, warning) and preparing associated metadata in advance, which significantly reduces the post-processing burden on the centralized server by presenting data in a ready-to-analyze format.
3Loss of information
If all sensor data is stored and transmitted, then data completeness is improved, but storage requirements and transmission costs increase
Solution Approach 1:
The patent applies local quality by differentiating the treatment of different data types based on their characteristics and requirements. Event message metadata is transmitted to the server for centralized analysis, while perception data (cameras, sensors) is retained locally in the vehicle's datastore where it can be accessed immediately. This differentiated approach ensures data completeness is maintained for both transmission and local access scenarios.
Solution Approach 2:
The patent introduces an intermediary mechanism where event messages serve as mediators between the local perception data and the centralized server. These event messages contain metadata that references and describes the perception data, enabling the server to understand and analyze events without requiring continuous transmission of the actual perception data, thus reducing transmission costs while maintaining analytical capability.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A computer-implemented method for monitoring driving experiences of an Automated Driving System (ADS) of a vehicle and related aspects are disclosed. The method comprises receiving, by an ADS monitoring layer in communication with the ADS of the vehicle and operating on a computing device of the vehicle, a query to monitor for an indication of an occurrence one or more event message trigger conditions and an occurrence one or more event recording trigger conditions by monitoring ADS data. Further, in response to fulfilment of the one or more event message trigger conditions, the method comprises transmitting, by the ADS monitoring layer, a first trigger signal to the ADS, storing, at a datastore of the ADS, an event message dataset comprising metadata associated with an event causing the fulfilment of the one or more event message trigger conditions, and transmitting the event message dataset to one or more remote servers. Further, in response to fulfilment of one or more event recording trigger conditions, the method comprises transmitting, by the ADS monitoring layer, a second trigger signal to the ADS, storing, at the datastore of the ADS, an event recording dataset comprising perception data associated with the event causing the fulfilment of the one or more event recording trigger conditions and/or ADS output data associated with the event causing the fulfilment of the one or more event recording trigger conditions, and transmitting the event recording dataset to the one or more remote servers. Here, the event recording dataset is larger than the event message dataset.