Autonomous Driving Platform Using Competitive Computing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional autonomous driving vehicles lack stability and accuracy due to variations in autonomous driving algorithms, leading to inconsistent performance in the same driving environments.
Innovation Solution
A method and server system utilizing competitive computing and information fusion, where a service server acquires and processes sensor data from multiple autonomous vehicles to generate credible autonomous driving source information, integrating it with circumstance-specific performance data to improve algorithm accuracy and stability, and enables safe autonomous driving by referring to self-verification scores and integrated information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional autonomous driving vehicles use their own autonomous driving algorithms based on deep learning, then they can perform autonomous driving, but different results are derived from different algorithms even in the same driving environment, leading to instability
Solution Approach 1:
The patent merges sensor data from multiple autonomous vehicles with different algorithms and combines their autonomous driving source information through information fusion. This integration allows the system to leverage diverse algorithmic perspectives while achieving stable and accurate autonomous driving decisions through competitive computing and data consolidation.
2Measurement precision
If autonomous vehicles use multiple different autonomous driving algorithms, then they can handle various driving scenarios, but accuracy and stability cannot be guaranteed due to inconsistent results
Solution Approach 1:
The patent implements a feedback mechanism where autonomous vehicles transmit their sensor data and autonomous driving source information to a service server. The server evaluates the credibility of different algorithms through competitive computing and provides feedback by selecting the most accurate algorithm for specific driving circumstances, thereby improving overall measurement precision while maintaining adaptability.
Solution Approach 2:
The system dynamically changes the parameter of algorithm selection based on driving circumstances. By evaluating circumstance-specific performance information and self-verification scores, the system adjusts which algorithm is used for specific scenarios, optimizing accuracy while preserving the versatility of having multiple algorithms available.
3Reliability
If autonomous vehicles rely solely on their own sensor data and algorithms, then they maintain operational independence, but credibility and safety of autonomous driving decisions are insufficient
Solution Approach 1:
The patent combines sensor data from multiple autonomous vehicles and integrates their autonomous driving source information through a service server. This merging of data resources enhances the credibility of driving decisions by leveraging collective information while the server manages the complexity of processing through systematic evaluation and selection.
4Measurement precision
If conventional autonomous driving systems use single-vehicle sensor data and algorithms, then they simplify data processing, but accuracy and stability of autonomous driving cannot be ensured
Solution Approach 1:
The patent introduces a service server as an intermediary that manages the complex task of processing sensor data from multiple vehicles. The server coordinates data collection, performs competitive computing evaluations, and facilitates information fusion, thereby achieving high measurement precision without significantly reducing data processing efficiency through centralized coordination.
Data Source
AI summary
A method for providing an autonomous driving service platform for autonomous vehicles by using a competitive computing and information fusion is provided. And the method includes steps of: (a) a service server acquiring individual sensor data and individual driving data through sensors installed on at least part of the autonomous vehicles including a subject vehicle; (b) the service server performing (i) a process of acquiring autonomous driving source information for the subject vehicle by inputting specific sensor data of specific autonomous vehicles among the autonomous vehicles and subject sensor data of the subject vehicle to data processing servers and (ii) a process of acquiring circumstance-specific performance information on the data processing servers from a circumstance-specific performance DB; and (c) the service server transmitting the autonomous driving source information and the circumstance-specific performance information to the subject vehicle, to thereby instruct the subject vehicle to perform the autonomous driving.