Autonomous Route Cost Mapping for Dynamic Obstacle Avoidance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing autonomous vehicles struggle to effectively detect and navigate around obstacles in limited outdoor areas, such as construction sites, where materials or other objects are placed on the route, compromising safety.
Innovation Solution
A route generation device for autonomous mobile bodies equipped with cameras, LIDARs, and radars to detect obstacles, generating a cost map that assigns lower values to the center of the route and higher values to edge zones and occupied areas, allowing the vehicle to avoid obstacles by traveling on low-cost grid cells and defining occupied cells as no-entry zones.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the autonomous mobile body uses a preset route for traveling, then the route planning is simple and efficient, but it cannot adapt to dynamic obstacles in real-time
Solution Approach 1:
The cost map is dynamically updated in real-time based on detection results from outside recognition devices. Grid cell cost values are continuously adjusted to reflect current obstacle positions, allowing the route generation device to adapt to changing environments while maintaining efficient pathfinding through the cost-based routing algorithm.
Solution Approach 2:
The system continuously receives detection results from outside recognition devices and uses this feedback to update the cost map. The updated cost map information is then fed back into the route generation process, creating a closed-loop system that adapts to real-time conditions while maintaining efficient route planning.
2Reliability
If the vehicle avoids all obstacles by taking alternative routes, then safety is improved, but travel time increases
Solution Approach 1:
The cost map applies different cost values to different grid cells based on their specific characteristics. Grid cells with obstacles have high cost values, while traversable areas have lower cost values. This local differentiation allows the system to avoid obstacles only where necessary while maintaining efficient routes through safe areas.
Solution Approach 2:
The system changes the cost parameter of grid cells dynamically based on obstacle detection. When an obstacle is detected, the cost value of affected grid cells is increased, which automatically influences route planning to avoid those areas. This parameter-based approach enables flexible adaptation without requiring complete route re规划.
3Measurement precision
If the cost map is updated frequently to reflect real-time obstacles, then obstacle detection accuracy is improved, but computational load increases
Solution Approach 1:
The system updates the cost map based on detection results from outside recognition devices, focusing computational resources on areas where obstacles are detected or where the vehicle is likely to travel. This partial updating approach maintains obstacle detection accuracy while reducing unnecessary computational overhead in clear areas.
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
Ensures safe navigation by avoiding obstacles, preventing vehicle departure from the lane, and maintaining safety even with dynamic obstacles, by generating routes that keep the vehicle at a safe distance from potential hazards.
Implementation Method 1
The autonomous mobile body (1) includes a plurality of LIDARs (12)
Implementation Method 2
a plurality of radars (13) that detect and recognize an outside of the vehicle body (10)
Data Source
AI summary
A route generation device for an autonomous mobile body 1 is configured to generate a route for the autonomous mobile body 1 based on a cost map of a way serving as the route for the autonomous mobile body 1, the cost map including a plurality of grid cells M1 to M9 into which the cost map is divided. In the cost map, a center of a road RD is assigned with a lower value of cost than an edge zone of the road RD. The route generation device includes a controller that is configured to update the cost map based on a result of detection by an outside recognition device, and generate a route based on the updated cost map.