Autonomous Vehicle Path Planning on Permissible Surfaces
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles often lack the ability to safely navigate and complete tasks without human assistance, especially in unfamiliar locations, and may inadvertently select routes through off-limit areas.
Innovation Solution
A computing system processes video data from a user device to identify permissible surfaces and develop a navigation path for the vehicle, allowing it to follow the demonstration path while making necessary variations to accommodate its capabilities and the delivery requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the autonomous vehicle follows the demonstration path directly, then it can reach the delivery location efficiently, but it may inadvertently select routes through off-limit areas
Solution Approach 1:
The patent introduces an intermediary processing system that receives the demonstration path from the user device, identifies permissible surfaces through image processing, and generates a modified navigation path. This intermediary layer acts as a mediator between the raw demonstration path and the vehicle's navigation system, ensuring that the final path both reaches the destination efficiently and avoids off-limit areas by constraining movement to identified permissible surfaces only.
2Extent of automation
If the vehicle navigates autonomously without human assistance, then operational independence is improved, but the ability to safely navigate unfamiliar locations deteriorates
Solution Approach 1:
The patent applies preliminary action by having the user device capture video data and identify permissible surfaces before the autonomous vehicle arrives at the unfamiliar location. The navigation path is pre-processed and validated against permissible surfaces in advance, so when the vehicle executes the path autonomously, it already has a verified safe route that accounts for the specific constraints of that location, thereby maintaining both autonomy and safety.
3Ease of operation
If the navigation path is adjusted to accommodate vehicle capabilities, then the vehicle can successfully complete delivery, but the path deviates from the original demonstration path
Solution Approach 1:
The patent applies local quality by making targeted, localized adjustments to the demonstration path only where necessary to accommodate vehicle capabilities and stay within permissible surfaces. Rather than completely deviating from the demonstration path, the system identifies specific segments that need modification and adjusts only those portions, preserving the overall fidelity and intent of the original demonstration path while ensuring practical executability by the autonomous vehicle.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Examples implementations relate to navigation path determination. An example method includes receiving, at a computing system, video data showing a demonstration path for navigating a location. The method further includes identifying, using the video data, a set of permissible surfaces at the location, wherein each permissible surface was traversed by the demonstration path. The method additionally includes determining a navigation path for a vehicle to follow at the location, wherein the navigation path includes a variation from the demonstration path such that the variation causes the vehicle to stay within one or more permissible surfaces from the set of permissible surfaces. The method also includes causing, by the computing system, the vehicle to follow the navigation path to navigate the location.