Scalable Data Fusion Architecture for Autonomous Driving
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data fusion architectures in autonomous driving systems are complex and costly, requiring high processing power and bandwidth, and lack flexibility to handle increasing sensor data and potential device failures.
Innovation Solution
A scalable data fusion method and system where both central devices and edge devices jointly manage data fusion, distributing workload and reducing complexity by allowing the number of edge devices to increase instead of enhancing the central device, ensuring system reliability through redundancy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If data fusion is performed centrally by a single device, then processing accuracy is improved, but device complexity and cost increase
Solution Approach 1:
The patent divides the centralized data fusion function into multiple edge devices, where each edge device performs local data fusion on sensor data from its associated sensors. This segmentation distributes the processing load across multiple devices rather than concentrating it in a single complex device, thereby maintaining processing accuracy while reducing individual device complexity and cost.
Solution Approach 2:
The patent introduces a hierarchical dimension to the data fusion architecture, with edge devices performing local fusion and a central device performing global fusion. This adds a spatial/organizational dimension to the processing architecture, allowing accuracy to be maintained through multiple fusion levels while distributing complexity across the hierarchy.
2Productivity
If processing power of the central device is enhanced to handle more sensor data, then data fusion capability is improved, but cost and complexity increase
Solution Approach 1:
The patent segments the data fusion workload by having edge devices perform initial data fusion locally before transmitting results to the central device. This segmentation allows the system to handle increased sensor data throughput without requiring a single powerful central device, instead using multiple moderately-capable edge devices that collectively provide the necessary processing capability.
Solution Approach 2:
The patent introduces edge devices as intermediary processing nodes between the sensors and the central device. These intermediaries perform preliminary data fusion and filtering, reducing the data burden on the central device and enabling the system to handle more sensor input without proportionally increasing central device complexity or cost.
3Productivity
If bandwidth is increased to transmit more sensor data, then data transmission capability is improved, but system cost increases
Solution Approach 1:
The patent applies preliminary data fusion at the edge devices before transmission to the central device. By fusing sensor data locally first, the volume of data that needs to be transmitted over the network is reduced, thereby maintaining high data transmission capability while avoiding the need for high-bandwidth (and high-cost) communication infrastructure.
4Measurement precision
If the system is designed to handle future HD sensors and higher definition data, then measurement precision is improved, but device complexity and bandwidth requirements increase
Solution Approach 1:
The patent segments the handling of high-definition sensor data across multiple edge devices, each capable of processing data from HD sensors locally. This segmentation allows the system to support future HD sensors and higher definition data without requiring a single overly complex central device, as the processing burden is distributed across the edge device network.
Data Source
AI summary
Provided are a scalable data fusion method and related products. The scalable data fusion method is applied in a central device and includes: receiving sensing data transmitted by each of M first edge devices, wherein M is an integer equal to or greater than 1; fusing the sensing data transmitted by each of the M first edge devices to obtain M pieces of fused data respectively corresponding to the M first edge devices; distributing the M pieces of fused data to the M first edge devices respectively; receiving object information transmitted by each of the M first edge devices, wherein the object information is obtained based on the fused data; and integrating the object information transmitted by each of the M first edge devices and construct surrounding information based on the integrated object information.