Vehicle Egospace Mapping for Static and Moving Obstacle Avoidance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems fail to efficiently and simultaneously represent and navigate around both static and moving obstacles for automatically controlled vehicles, leading to challenges in collision avoidance and trajectory planning, especially in complex three-dimensional environments.
Innovation Solution
A combined egospace representation is used to compactly integrate static and moving obstacle data, allowing for the efficient generation of collision-free trajectories by merging moving and static obstacle information into a single data structure, enabling vehicles to avoid collisions and intercept targets while navigating through environments with both types of obstacles.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If separate representation methods are used for static and moving obstacles, then each obstacle type can be handled with specialized algorithms, but the system complexity increases and simultaneous navigation becomes difficult
Solution Approach 1:
The patent merges the representation of static and moving obstacles into a unified egospace data structure. Both obstacle types are encoded using the same coordinate system (radial distance, azimuth, elevation) and stored together in a single three-dimensional array, allowing simultaneous navigation without requiring separate processing systems.
Solution Approach 2:
The egospace representation serves multiple functions simultaneously: it represents both static and moving obstacles, enables collision detection, supports trajectory planning, and provides a common framework for different obstacle types. This universal representation eliminates the need for separate specialized algorithms.
2Measurement precision
If complex three-dimensional environment data is processed in detail, then navigation accuracy improves, but computational time and processing resources increase
Solution Approach 1:
The patent transforms three-dimensional obstacle data into a different dimensional representation (egospace with radial, azimuth, and elevation coordinates). This dimensional transformation organizes the data in a way that preserves navigation accuracy while enabling more efficient computational processing through structured array access patterns.
Solution Approach 2:
The representation parameters are changed from Cartesian coordinates to spherical-like coordinates (radial distance, azimuth, elevation). This parameter transformation maintains the precision needed for navigation while optimizing the data structure for computational efficiency in trajectory planning and collision detection.
3Measurement precision
If high-resolution obstacle representation is used, then collision detection accuracy improves, but memory requirements and data storage increase
Solution Approach 1:
The patent segments the three-dimensional space into discrete angular bins (azimuth and elevation) and radial distances, creating a structured grid representation. This segmentation maintains collision detection accuracy by preserving spatial resolution while organizing data in a compact array format that optimizes memory usage.
Solution Approach 2:
The egospace representation nests multiple levels of spatial information within a hierarchical structure: radial distance layers contain angular information, which in turn contain obstacle presence data. This nested organization preserves detailed collision detection capability while compressing the overall data structure to reduce memory requirements.
Data Source
AI summary
A method and system provide the ability to automatically control a vehicle to avoid obstacle collision. Range data of a real-world scene (including depth data to static objects) is acquired. Positions and velocities of moving objects are acquired. The range data is combined into an egospace representation for pixels in egospace that is specified with respect to a radially aligned coordinate system. An apparent size of the static objects is expanding, in the egospace representation, based on a dimension of the vehicle. A speed of the vehicle is specified. A velocity obstacle corresponding to the moving objects is constructed. A mask is created in the coordinate system and identifies candidate radial paths that will result in a collision between the vehicle and the moving objects. The mask is combined with the egospace representation that is then used to determine a path for the vehicle.