Vehicle Surrounding Intelligence via Network-Based Sensor Fusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in making safe decisions due to unreliable sensor data in challenging environments, lacking information on nearby vehicles' behavior and intentions, which can lead to traffic issues or collisions.
Innovation Solution
Implement a centralized approach where a network entity, such as the LMF, provides surrounding intelligence information by analyzing and predicting the locations and intentions of nearby vehicles, refining sensor data, and categorizing vehicles based on their capabilities to enhance decision-making.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If vehicles rely on local sensor data (cameras, RADARs, LiDARs) for surrounding information, then the vehicle can obtain real-time local environmental data, but the reliability of this information deteriorates under challenging conditions such as poor illumination, weather, sensor blockage, and limited detection range
Solution Approach 1:
The patent introduces a network entity (LMF) as an intermediary that collects, processes, and fuses sensor data from multiple vehicles and external sources. This mediator aggregates information that individual vehicles cannot obtain alone, providing reliable surrounding intelligence even when local sensors fail due to weather, blockage, or illumination issues
Solution Approach 2:
The patent combines data from multiple sources including local sensors, remote sensors, map data, and historical traffic information through data fusion at the network entity. This merging of diverse information sources compensates for the limitations of individual sensor types and creates a more complete and reliable picture of the surrounding environment
2Loss of information
If vehicles use only local sensor data for decision-making, then the system remains decentralized and simple, but the vehicle lacks predictive information about nearby vehicles' intentions and future positions
Solution Approach 1:
The LMF acts as a centralized intermediary that performs complex analytics on aggregated data from multiple vehicles. It predicts future positions, determines intentions, and processes this information centrally, then provides the results back to individual vehicles. This approach maintains simplicity at the vehicle level while enabling sophisticated predictive capabilities through centralized processing
Solution Approach 2:
The network entity performs preliminary analysis and prediction of vehicle behavior, intentions, and future positions before the vehicles need to make decisions. By pre-processing this information centrally and making it available to vehicles, the system provides predictive insights without requiring each vehicle to perform complex real-time analytics
3Reliability
If vehicles make decisions based on current position only, then the decision-making process is simple and fast, but the vehicle cannot predict future collisions or assess safety of maneuvers
Solution Approach 1:
The LMF performs preliminary calculations of future vehicle positions, collision risks, and safe maneuver assessments in advance, before vehicles need to make decisions. This pre-computed safety information is provided to vehicles so they can make decisions based on already-analyzed data rather than performing time-consuming real-time calculations
Solution Approach 2:
The system provides feedback to vehicles about predicted collision risks, safe velocities, and recommended maneuvers based on continuous monitoring and analysis of surrounding traffic. This feedback loop enables vehicles to make safer decisions by incorporating predicted future states and collision assessments into their decision-making process
Data Source
Figure 1
Figure 2
Figure 3
AI summary
There are provided measures for provision of vehicle's surrounding intelligence information. Such measures exemplarily comprise, at a network side node of a cellular system, receiving, from a first mobile entity, a request for combined spatiotemporal surroundings characteristics related to surroundings of said first mobile entity, generating said combined spatiotemporal surroundings characteristics related to said surroundings of said first mobile entity based on a position of said first mobile entity, and transmitting, towards said first mobile entity, said combined spatiotemporal surroundings characteristics related to said surroundings of said first mobile entity, wherein said combined spatiotemporal surroundings characteristics related to said surroundings of said first mobile entity include location information of at least one second mobile entity in the surroundings of said first mobile entity.