Autonomous Vehicle Control Model Updating Through Targeted Sensor Tasks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for updating control models for autonomous vehicles are inefficient and costly, as they require extensive data collection in various driving situations, which can be time-consuming and resource-intensive.
Innovation Solution
A method and system where a central control unit generates a recording order for mobile units to collect specific sensor data, which is then used to generate an updated control model, allowing for targeted data collection and optimization through exploratory learning and cooperative acquisition, enabling the identification and optimization of weaknesses in automatic driving functions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If extensive data collection is performed to improve control model accuracy, then the quality of automatic driving control is improved, but the time and resources required for data collection increase significantly
Solution Approach 1:
The system performs preliminary analysis to identify specific weaknesses and deficiencies in the control model before data collection. By determining exactly what aspects need improvement (e.g., specific driving situations, sensor types, or control parameters), the system can pre-define targeted data acquisition tasks that focus only on collecting relevant data for those specific weaknesses, rather than collecting extensive general data.
Solution Approach 2:
Instead of uniformly collecting data across all driving situations and sensor types, the system applies local quality by tailoring data collection to specific local needs. The central control unit generates different data acquisition tasks for different mobile units based on their specific weaknesses, and each mobile unit collects data only for the situations and parameters where improvements are needed, rather than collecting all possible data uniformly.
2Adaptability or versatility
If comprehensive data collection covers all driving situations to improve model robustness, then the adaptability of the control model is improved, but the complexity and cost of the data acquisition system increase
Solution Approach 1:
The system performs preliminary analysis to identify specific weaknesses and deficiencies in the control model before data collection. By determining exactly what aspects need improvement (e.g., specific driving situations, sensor types, or control parameters), the system can pre-define targeted data acquisition tasks that focus only on collecting relevant data for those specific weaknesses, rather than collecting extensive general data.
Solution Approach 2:
Instead of uniformly collecting data across all driving situations and sensor types, the system applies local quality by tailoring data collection to specific local needs. The central control unit generates different data acquisition tasks for different mobile units based on their specific weaknesses, and each mobile unit collects data only for the situations and parameters where improvements are needed, rather than collecting all possible data uniformly.
3Reliability
If test drivers manually collect data in problematic situations to improve automated driving, then the quality of specific edge cases is improved, but the cost and time consumption increase significantly
Solution Approach 1:
The system enables automated data collection where mobile units equipped with sensors automatically collect the required sensor data sets based on generated data acquisition tasks. This eliminates the need for manual test driver intervention, as the mobile units themselves perform the data collection autonomously by executing predefined tasks that target specific weaknesses in the control model.
Solution Approach 2:
The system implements a feedback loop where the central control unit receives sensor data from mobile units, analyzes the results, and generates updated data acquisition tasks based on identified weaknesses. This automated feedback mechanism continuously improves the control model without requiring manual test driver intervention, as the system self-corrects by generating new tasks based on previous performance analysis.
Data Source
Figure 1A~1B
Figure 2~3
AI summary
The invention relates to a method and a corresponding system for updating a control model for an automatic control of at least one mobile unit (1). A central control unit (2) generates a detection task and transmits same to the mobile unit (1). The mobile unit (1) comprises sensors (3a; 3b, 3c; 3d), and the detection task comprises conditions for detecting sensor data sets by means of the sensors (3a; 3b, 3c; 3d). The mobile unit (1) detects the sensor data sets by means of the sensors (3a; 3b, 3c; 3d) using the detection task, generates transmission data using the detected sensor data sets, and transmits the transmission data to the central control unit (2). The central control unit (2) receives the transmission data and generates an updated control model using the received transmission data.