Drivable Surface Estimation Using Flatness and Texture Fusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The validation and verification of autonomous driving systems are costly and time-consuming due to the combinatorial explosion of possible situations, necessitating a high probability of failure-free operation, which existing methods struggle to efficiently address.
Innovation Solution
A method and system for estimating a drivable space using independent algorithms for surface flatness and texture, utilizing sensors like LIDAR and cameras, to determine the presence of drivable space by fusing and time-filtering sensor data, reducing validation effort through redundant estimation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional testing methods are used to validate autonomous systems, then safety and reliability requirements can be met, but the validation process becomes extremely costly and time-consuming due to combinatorial explosion of possible situations
Solution Approach 1:
The patent segments the complex validation problem into two independent sub-problems: surface flatness estimation and surface texture estimation. By dividing the drivable surface detection task into separate algorithmic components, the system can validate each segment independently rather than testing all possible combinations of surface conditions, thereby reducing validation complexity and time while maintaining reliability
Solution Approach 2:
The patent introduces an intermediary computational layer that processes sensor data through separate estimation algorithms for flatness and texture. This intermediary processing stage transforms raw sensor data into structured surface characteristics that can be validated independently, serving as a mediator between raw data and final drivable surface determination
2Reliability
If multiple independent algorithms are used for surface estimation, then validation effort is reduced through independent verification, but device complexity increases
Solution Approach 1:
The system segments the estimation function into distinct independent algorithms: one for surface flatness and another for surface texture. Each algorithm has a specific, narrow function that can be validated separately, reducing overall validation effort despite the presence of multiple algorithms. The segmentation allows each component to be simpler and more verifiable
Solution Approach 2:
The patent uses multiple independent algorithms that essentially copy the same estimation function from different perspectives (flatness estimation and texture estimation). This redundancy provides independent verification paths without requiring a single complex algorithm, as each copy focuses on a specific aspect of surface characterization
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Reduces the validation effort by several orders of magnitude, ensuring reliable and safe autonomous driving by independently estimating surface flatness and texture, thereby meeting stringent safety requirements.
Implementation Method 1
utilizing sensors like LIDAR and cameras
Data Source
AI summary
The present disclosure relates to a control device and method for estimating a drivable space in a surrounding environment of a vehicle. In particular, the disclosure relates to a “free space estimation” solution with reduced validation effort. The control device includes one or more processors adapted to obtain sensor data, determine a surface flatness MF and a surface texture MR of at least one portion of the surrounding environment by means of two independent algorithms, and defining a drivable space based on obtained values.