Vehicle Localization Across Mixed Sensor Coverage Areas
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous and remotely controlled vehicles face challenges in accurately determining their location across areas with different types of sensor coverage, leading to potential safety risks due to inconsistent and discontinuous positioning.
Innovation Solution
A method and system for localizing vehicles by calibrating sensor data from multiple types of sensors with different coordinate systems into a common positioning system, using transforms to ensure continuous and consistent vehicle localization, even when transitioning between sensor coverage areas.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If multiple different types of sensors with different positioning coordinate systems are used in different sensor coverage areas, then the coverage area can be expanded and more areas can be monitored, but the positioning consistency and continuity deteriorate when transitioning between different sensor coverage areas
Solution Approach 1:
The patent introduces a coordinate transformation system as an intermediary that converts positioning data from different sensor coordinate systems into a common reference coordinate system. This mediator enables seamless integration of GPS, inertial sensors, and other positioning systems by transforming their respective coordinate data into a unified framework, thereby maintaining positioning consistency across different sensor coverage areas.
Solution Approach 2:
The patent dynamically adjusts transformation parameters (such as rotation matrices, translation vectors, and scaling factors) based on the current sensor coverage area and transition state. By changing these parameters adaptively during transitions between different sensor zones, the system maintains accurate and continuous positioning despite using different sensor types with different coordinate systems.
2Adaptability or versatility
If different types of sensors are used in different sub-areas for vehicle localization, then the adaptability to various environments is improved, but the device complexity increases due to coordinate system mapping and transformation
Solution Approach 1:
The patent divides the operational area into distinct sensor coverage zones, each optimized for specific environmental conditions (e.g., GPS-rich open areas, inertial-sensor-dependent tunnel areas). By segmenting the environment and assigning appropriate sensor types to each zone, the system achieves environmental adaptability while managing complexity through localized sensor deployment rather than universal multi-sensor coverage.
Solution Approach 2:
The patent implements a universal coordinate transformation framework that can handle multiple sensor types and coordinate systems through a single unified approach. This multi-functional transformation system processes data from GPS, inertial sensors, and other positioning systems using the same mathematical framework, reducing overall system complexity despite the diversity of sensor inputs.
3Loss of time
If sensor data from different coordinate systems is directly used without calibration, then the processing time is reduced, but the localization accuracy deteriorates due to global measurement errors
Solution Approach 1:
The patent performs coordinate system calibration and transformation parameter determination in advance, before actual positioning operations. By pre-computing transformation matrices and calibration factors during system initialization or offline setup, the system eliminates the need for real-time complex calculations, thus maintaining both high processing speed and high localization accuracy during vehicle operation.
Data Source
AI summary
A method of localizing a vehicle in an area of vehicle operation includes receiving sensor data at the vehicle from a plurality of different types of sensors positioned in the area, at least two of the different types of sensors providing location data using different positioning coordinate systems, the different positioning coordinate systems are non-linear with respect to each other; mapping the different positioning coordinate systems to a common coordinate system, including splitting the area into a plurality of subareas based on sensor coverage in each subarea and assigning a transform to be used for each subarea; automatically calibrating received sensor data associated with different positioning coordinate systems to form sensor data associated with a common positioning coordinate system in each subarea; and localizing the vehicle using the calibrated sensor data associated with the common positioning coordinate system in a subarea of interest in the area, provided by the different types of sensors to provide the sensor data associated with the common positioning coordinate system.