Object Uncertainty Heatmaps for Drivable Area Collision Checks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face computational challenges in safely navigating through environments due to the high cost of maintaining and updating uncertainty associated with predicted object behavior, especially in crowded areas where millions of operations are required per second.
Innovation Solution
The system fuses uncertainty data of detected objects into a single uncertainty model representing the physical environment, using heatmaps and symmetric, rotationally invariant disks to reduce computational complexity and simplify collision checks, and represents objects as enlarged regions to account for uncertainty and potential trajectories.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the system maintains and updates uncertainty associated with each detected object individually, then the accuracy of collision detection is improved, but the computational cost increases significantly requiring millions of operations per second
Solution Approach 1:
The patent merges individual uncertainty data from multiple detected objects into a single aggregated uncertainty model representing the entire environment. Instead of maintaining separate uncertainty calculations for each object, the system combines them into one unified representation that can be processed efficiently while still providing accurate collision detection capabilities.
Solution Approach 2:
The uncertainty model serves multiple functions simultaneously: it represents environmental uncertainty, enables collision detection, supports trajectory prediction, and provides a common framework for various navigation decisions. This multi-functional approach eliminates the need for separate computational systems for each function.
2Reliability
If the system represents objects with detailed uncertainty regions, then the safety of navigation is improved, but the device complexity increases
Solution Approach 1:
The patent transforms the representation of uncertainty by changing parameters from individual object-level detailed models to a unified environmental model with aggregated uncertainty parameters. This parameter transformation maintains the necessary safety information while reducing the overall complexity of the system.
3Measurement precision
If the system performs comprehensive uncertainty updates for all objects in crowded areas, then the accuracy of drivable area determination is improved, but the processing time increases
Solution Approach 1:
The system combines uncertainty information from all objects into a single aggregated model that represents the overall environmental uncertainty. This merging allows the system to determine drivable areas by querying the unified model rather than performing separate uncertainty calculations for each object, dramatically reducing processing time while maintaining accuracy.
Data Source
AI summary
Techniques for representing sensor data and predicted behavior of various objects in an environment are described herein. For example, an autonomous vehicle can represent prediction probabilities as an uncertainty model that may be used to detect potential collisions, define a safe operational zone or drivable area, and to make operational decisions in a computationally efficient manner. The uncertainty model may represent a probability that regions within the environment are occupied using a heat map type approach in which various intensities of the heat map represent a likelihood of a corresponding physical region being occupied at a given point in time.