V2X Sensor Information Sharing for Occluded Object Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Self-driving vehicles and other driving assist mechanisms face challenges in detecting all objects due to occlusions and adverse weather or lighting conditions, and existing communication systems are inefficient in sharing sensor information among vehicles, leading to redundant data transmission.
Innovation Solution
A request-response communication scheme using V2X (Vehicle-to-Anything) communication allows vehicles to share sensor information efficiently by determining which vehicles should transmit data and when, reducing redundant information and improving data transfer efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If vehicles continuously broadcast sensor information, then all vehicles can access sensor data, but data redundancy increases and communication efficiency decreases
Solution Approach 1:
The system performs preliminary actions by having vehicles publish sensor information periodically based on predicted occlusion events, rather than continuously. The requestor vehicle predicts when occlusion events will occur and requests information in advance, allowing the responder to prepare and transmit data at optimal times, reducing redundancy while ensuring availability.
Solution Approach 2:
The system implements feedback through request-response communication where the requestor vehicle sends requests based on its sensing capabilities and occlusion predictions, and the responder vehicle adjusts its transmission based on received requests. This feedback loop optimizes data sharing by only transmitting information when and where it is needed, reducing redundancy.
2Reliability
If vehicles request sensor information from all nearby vehicles, then detection coverage improves, but communication overhead increases
Solution Approach 1:
The system applies local quality by tailoring the information sharing to specific local needs. Each requestor vehicle identifies which responder vehicles are most relevant based on their relative positions and sensing capabilities, and requests only the specific sensor information needed for objects in occluded regions. This localized approach improves detection reliability without requiring communication with all nearby vehicles.
Solution Approach 2:
The system segments the sensor information sharing process into discrete request-response transactions rather than continuous broadcast. Each request targets specific information from specific vehicles based on occlusion analysis, breaking down the complex task of obtaining comprehensive sensor data into manageable, targeted queries.
3Loss of information
If vehicles transmit all sensor data, then information completeness improves, but data redundancy increases
Solution Approach 1:
The system extracts only the necessary sensor information from the total available data. The requestor vehicle identifies specific objects and regions of interest based on its sensing capabilities and occlusion predictions, then requests only the sensor data relevant to those objects. This extraction approach maintains information completeness for critical areas while reducing overall data volume by excluding redundant information from non-occluded regions.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The invention relates to communication for an apparatus of a first object, comprising, at the apparatus: determining that a second object comprising a second apparatus is within at least a threshold proximity from at least one sensor of the first object and coupled to the apparatus; determining a set of detected objects using, at least in-part, the at least one sensor; determining a subset of the set of detected objects that are presently occluded or are expected to be occluded from sensor detection from the second object; and transmitting a message to the second apparatus, the message comprising information corresponding to at least one detected object of the subset of the set of detected objects. rt, on the sensor information of the received message.