Occupancy Grid Control for Autonomous Vehicle Surface Discontinuities
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles (AVs) face challenges in navigating complex terrains with discontinuous surface features (SDSFs) due to the limitations of existing sensor technologies, which struggle to accurately identify and traverse such features in real-time, affecting their ability to adapt vehicle configuration and ensure safe passage.
Innovation Solution
The system combines advanced sensor data processing with real-time vehicle configuration changes, using a multi-part model for SDSF identification and traversal, incorporating long-range sensors, short-range sensors, and a perception subsystem to create an occupancy grid that informs path planning and vehicle configuration adjustments, enabling precise navigation over SDSFs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If long-range sensors and short-range sensors are used for navigation and obstacle detection, then the AV can sense and avoid obstacles, but the AV struggles to accurately identify and traverse discontinuous surface features in real-time
Solution Approach 1:
The patent segments the surface detection task into multiple components: long-range sensors detect potential SDSFs at distances of 4-100 meters, short-range sensors verify and characterize them at 5 meters, and the occupancy grid system processes this segmented information hierarchically. This segmentation allows the system to maintain both detection accuracy and real-time reliability by distributing the computational load across multiple sensor levels.
2Adaptability or versatility
If the AV reconfigures its physical configuration in real-time to traverse SDSFs, then the AV can adapt to complex terrains, but the system complexity increases
Solution Approach 1:
The occupancy grid system performs preliminary action by pre-processing sensor data to identify and map SDSFs before the AV encounters them. The system creates a probabilistic representation of the environment that anticipates terrain challenges, allowing the vehicle to plan configuration changes in advance rather than reacting in real-time, thus reducing control system complexity while maintaining adaptability.
Solution Approach 2:
The patent implements dynamics by making the AV's physical configuration adjustable in real-time based on detected terrain. The vehicle can change wheel configurations, suspension settings, or body orientation dynamically as it traverses different surface features. This dynamic adaptability is managed through the occupancy grid system that provides real-time terrain information to the control system.
3Measurement precision
If the AV uses occupancy grid with probability representation for path planning, then the decision-making accuracy improves, but the computational processing requirements increase
Solution Approach 1:
The patent applies parameter changes by using logodds transformation to represent occupancy probabilities. This mathematical transformation changes the parameter representation from raw probabilities to logodds values, which improve numerical stability at boundaries (near 0 and 1) and enable more efficient computational operations. This parameter change maintains high decision-making accuracy while reducing computational energy consumption through more stable numerical operations.
Data Source
AI summary
An autonomous vehicle having sensors advantageously varied in capabilities, advantageously positioned, and advantageously impervious to environmental conditions. A system executing on the autonomous vehicle that can receive a map including, for example, substantially discontinuous surface features along with data from the sensors, create an occupancy grid based upon the map and the data, and change the configuration of the autonomous vehicle based upon the type of surface on which the autonomous vehicle navigates. The device can safely navigate surfaces and surface features, including traversing discontinuous surfaces and other obstacles.