ADS Perception Model Updating Using a Joint World View
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The development and verification of Autonomous Driving Systems (ADS) are hindered by the need for extensive training data and the challenge of managing immense data generated, which is costly and time-consuming, especially in ensuring safety and performance improvements.
Innovation Solution
An assessment system that combines world view data from an ego-vehicle's perception module with data from surrounding vehicles to create a joint world view, allowing for the evaluation and updating of perception models by matching perception data to the joint world view, thereby improving the accuracy and efficiency of perception features in ADS.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If extensive training data is collected and annotated for supervised training of perception models, then the performance and accuracy of ADS perception features is improved, but the cost and time required for data collection and processing increases significantly
Solution Approach 1:
The patent combines world view data from multiple vehicles (ego-vehicle and other vehicles) to create a joint world view. This merging of data sources allows the perception model to be trained on aggregated data from multiple vehicles simultaneously, improving model accuracy without requiring each individual vehicle to collect and process extensive data independently, thus reducing the time and computational resources needed for data processing.
Solution Approach 2:
The system performs preliminary matching of perception data to the joint world view before detailed evaluation. By pre-aligning and organizing data from multiple vehicles into a coordinated reference system, the system reduces the subsequent processing burden and accelerates the training pipeline, addressing the time loss issue while maintaining model accuracy.
2Measurement precision
If extensive training data is collected and annotated for supervised training of perception models, then the performance and accuracy of ADS perception features is improved, but the system cost increases significantly
Solution Approach 1:
By merging data from multiple vehicles into a joint world view, the system creates a shared training resource that benefits all participating vehicles. This eliminates the need for each vehicle to independently collect and annotate extensive training data, significantly reducing the overall development cost while maintaining or improving perception model accuracy through the aggregated diverse dataset.
Solution Approach 2:
The system creates a virtual joint world view that copies and integrates perception data from multiple vehicles. This virtual representation allows the perception model to learn from diverse scenarios without requiring physical duplication of test environments or manual annotation efforts for each vehicle, reducing development costs while improving model generalization.
3Productivity
If world view data from multiple vehicles is combined to create a joint world view, then the efficiency and cost-effectiveness of perception model development is improved, but the device complexity and data management requirements increase
Solution Approach 1:
The patent introduces a joint world view as an intermediary structure that mediates between raw perception data from multiple vehicles and the perception model training process. This intermediary layer handles the complexity of data integration, coordinate transformation, and synchronization, simplifying the overall system architecture while enabling efficient use of multi-vehicle data for model development.
Solution Approach 2:
The system replaces complex manual data management processes with automated computational methods for creating and maintaining the joint world view. By using algorithmic approaches for data fusion, coordinate system transformation, and temporal synchronization, the system reduces manual intervention requirements while managing the complexity of multi-vehicle data integration, thereby improving development efficiency.
Data Source
AI summary
An assessment system for performance evaluation and updating of a PMUD of an ego-vehicle. The assessment system obtains world view data from a perception module configured to generate the world view data based on sensor data obtained from vehicle-mounted sensors; obtains other world view data generated by another perception module; forms a joint world view by matching the world view data the other world view data; and obtains perception data based on a perception model and sensor data obtained from one or more vehicle-mounted sensors. The assessment system further matches the perception data to the formed joint world view; evaluates the obtained perception data in reference to the joint world view to determine an estimation deviation in an identified match between the perceptive parameter of the perception data and a corresponding perceptive parameter in the joint world view; and updates parameters of the perception model based on the estimation deviation.