Vehicle Perception Sharing via Edge Fusion for Blind Spot Coverage
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems fail to robustly monitor high-density real-time traffic situations, which hampers smooth traffic flow and effective response of emergency vehicles, particularly in areas where advanced driver assistance systems are limited by processing latency and blind spots caused by intervening vehicles.
Innovation Solution
A vehicle-to-everything (V2X) communication system with a multi-access edge computing cluster and roadside units that collects and fuses real-time data from multiple vehicles to predict object motion, identify locations, and communicate fused perception results, enabling blind area monitoring and reducing processing latency through standardized APIs and efficient software architecture.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If vehicles use individual sensor systems to monitor traffic, then each vehicle can detect objects in its field of view, but blind spots are created by intervening vehicles and processing latency increases in high-density traffic
Solution Approach 1:
The patent merges sensor data from multiple vehicles to create a collective perception system. The server aggregates detection results from several vehicles, allowing the group to monitor areas that would be blind spots for individual vehicles. This combining of information sources eliminates blind spots caused by intervening vehicles while maintaining reliable traffic monitoring in high-density environments.
Solution Approach 2:
The patent introduces a server as an intermediary that collects, processes, and distributes perception data among vehicles. This mediator coordinates the sharing of sensor information, enabling vehicles to access data from other vehicles' sensors and thereby overcoming individual blind spots without direct vehicle-to-vehicle communication complexity.
2Productivity
If vehicles process sensor data individually, then each vehicle maintains independent decision-making, but processing latency increases in high-density traffic situations
Solution Approach 1:
The patent moves the data processing function from the individual vehicle dimension to a centralized server dimension. By relocating the computational workload to a dedicated server that handles perception data from multiple vehicles, the system reduces the processing burden on each vehicle and enables more efficient handling of high-density traffic scenarios with lower latency.
Solution Approach 2:
The patent extracts the complex perception processing function from individual vehicles and consolidates it on a separate server. This separation allows vehicles to focus on execution while the server handles the computationally intensive tasks of aggregating and analyzing sensor data from multiple sources, thereby reducing overall system latency.
3Reliability
If a centralized system collects data from all vehicles, then comprehensive traffic monitoring is achieved, but system complexity and communication overhead increase
Solution Approach 1:
The patent creates a universal server platform that handles multiple functions: collecting sensor data from various vehicle types, processing different sensor formats, generating perception results, and distributing information back to vehicles. This multi-functional design consolidates complexity into a single platform that can serve diverse vehicles, reducing overall system complexity compared to individual vehicle systems.
Data Source
AI summary
A method for perception-sharing between similarly-situated vehicles that are traveling on a portion of a roadway that is equipped with an intelligent vehicle highway system is described, and includes executing, in a multi-access edge computing cluster in communication with a roadside unit disposed to monitor a roadway, an application-layer routine. The application-layer routine includes collecting real-time data associated with a plurality of objects from each of the similarly-situated vehicles, predicting motion of each of the plurality of objects based upon the real-time data, object-matching the motion of each of the plurality of objects, and executing fusion of the plurality of objects based upon the object-matching of the motion of each of the plurality of objects. Locations of the similarly-situated vehicles traveling on the roadway are identified based upon the fusion of the plurality of objects. The locations are communicated to one of the similarly-situated vehicles.