Cooperative Vehicle Sensing for Redundant Object Perception
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicle systems face reliability challenges due to errors in sensor interpretation, particularly in crowded conditions, and existing methods to increase redundancy result in increased system complexity and cost, making them prone to malfunctions.
Innovation Solution
A vehicle system that utilizes cooperative sensing and decision-making by sharing sensory data between interconnected vehicles to improve redundancy and accuracy, minimizing signal noise by delaying coordinate transformations to the receiving side and using probabilistic filters to remove anomalies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If redundancy is increased by using multiple independent sources for decision-making, then reliability is improved, but system complexity and cost increase
Solution Approach 1:
The patent merges sensing capabilities across multiple vehicles by having them share sensory data through communication systems. Instead of each vehicle independently implementing redundant sensing hardware, the system combines data from multiple vehicles' sensors to achieve redundancy, thereby improving reliability without increasing individual vehicle system complexity
Solution Approach 2:
The sensory data collected by each vehicle serves multiple functions: it is used for the collecting vehicle's own perception and is simultaneously shared to provide redundancy for other vehicles in the network. This multi-functional use of data reduces the need for dedicated redundant components in each vehicle
2Reliability
If more components are added to increase redundancy, then reliability is improved, but the system becomes more prone to malfunctions
Solution Approach 1:
The system creates virtual copies of sensing capabilities by sharing data from other vehicles' sensors. Instead of physically duplicating sensor hardware (which would increase malfunction risk), the system copies sensory information through communication networks, achieving redundancy without additional physical components that could fail
3Ease of operation
If coordinate transformations are performed early in the processing chain, then data integration is simplified, but signal noise increases
Solution Approach 1:
The system performs preliminary data collection and sharing in the vehicles' respective local coordinate systems before any transformations are applied. By delaying coordinate transformations until the final fusion stage at the receiving vehicle, the system avoids accumulating transformation errors and noise that would occur with early transformations, while still enabling data integration from multiple sources
Data Source
AI summary
Embodiments of the invention relate to a vehicle system for an autonomous or semi-autonomous motor vehicle, and a perception system for managing a plurality of interconnected motor vehicles. More specifically, the vehicle system has a control unit for retrieving a relative position of an object in a surrounding environment of the vehicle by means of a sensory system and for retrieving a heading of the motor vehicle. Furthermore, the control unit is configured to receive data from a different motor vehicle comprising the heading of the different motor vehicle and the position of the object relative to the different motor vehicle. The control unit is further configured to translate this external measurement to a local coordinate system in order to add redundancy to the measurement (and thereby accuracy) by utilizing a distributed system approach and therefore reduce the need for adding cost and complexity to each vehicle system in order to achieve increased accuracy.