Autonomous Vehicle Control Device Modular Data Weighting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing control devices for autonomous vehicles face challenges in achieving robustness and flexibility due to complex algorithms and sensor compromises, which affect the quality of autonomous operation, especially in diverse situations and environments.
Innovation Solution
A modular software environment with decision-making modules that evaluate and weight data from multiple sensors and algorithms, ensuring data quality and redundancy, allowing for flexible selection of the best data sources based on criteria such as usability and quality, thereby enhancing robustness and flexibility.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If platform and task-specific computer programs are developed to cover diverse situations and environments, then the adaptability and versatility of the control device is improved, but the device complexity and algorithm complexity increase significantly
Solution Approach 1:
The control device is segmented into multiple independent data sources (sensors and computer programs) that can be selectively activated. Each data source operates independently and provides specific data types, allowing the system to handle diverse situations without requiring a single complex algorithm to cover all cases.
Solution Approach 2:
The system dynamically selects and weights data sources based on current operational conditions, task requirements, and data quality metrics. The weighting factors are adjusted in real-time to optimize performance for the current situation, enabling adaptability without permanent complexity increases.
2Reliability
If multiple sensors and computer programs are used to provide redundant data sources, then the reliability and robustness of the control device is improved, but the device complexity and data processing requirements increase
Solution Approach 1:
Multiple data sources are merged through a weighted evaluation mechanism that combines their outputs into a unified data stream for downstream computer programs. The merging process uses quality weights to harmonize data from different sources, reducing the need for complex individual processing of each source while maintaining reliability.
Solution Approach 2:
A decision-making module acts as an intermediary between multiple data sources and downstream computer programs. This intermediary evaluates data quality, assigns weights, and selects optimal data combinations, simplifying the architecture by centralizing the complexity in a dedicated module rather than distributing it throughout the entire system.
3Measurement precision
If data from multiple data sources is evaluated and weighted based on quality criteria, then the measurement precision and data quality are improved, but the loss of time for data evaluation increases
Solution Approach 1:
Quality criteria and weighting factors are predetermined and stored for different data sources and operational conditions. When a data source becomes available, its pre-defined quality metrics are immediately applied without requiring complex real-time analysis, thus maintaining high data quality while minimizing evaluation time.
Solution Approach 2:
The system changes parameters (weighting factors) based on pre-established quality criteria rather than performing complex real-time optimization. This allows rapid data evaluation by switching between pre-computed weight sets corresponding to different operational modes and data source qualities.
Data Source
Figure 1~2
Figure 3
AI summary
The method involves performing intercommunicating computer programs of software package for implementing autonomous operation function. The transmitted data is evaluated by portion of computer programs of sensors (2,3) and/or different computer program. The identical data is processed by different algorithm investigating computer programs and/or designed to determine similar sensor data as different data sources. The data from sensors are sent to computer program unit as input data through decision modules by using algorithms and/or sensors described criterion.