Topological Map Updating for Traversable Robot Navigation Paths
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing robotic navigation systems face inaccuracies in topological maps due to odometry drift and localization errors, leading to incorrect graph topologies and non-traversable edges, which hinder efficient path planning and localization.
Innovation Solution
A method and system for verifying and updating topological maps by identifying candidate alternate edges using sensor data, confirming traversability, and optimizing the map to ensure consistency and accuracy, utilizing algorithms like iterative closest point and sparse nonlinear optimization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional odometry-based topological mapping is used, then the robot can build a navigation map, but the map contains inaccuracies due to odometry drift and localization errors
Solution Approach 1:
The system uses sensor data (LIDAR, cameras) to continuously verify and correct the topological map against actual environmental features. Loop closure detection provides feedback when the robot returns to previously visited locations, allowing the system to detect and correct accumulated odometry errors. This feedback mechanism ensures map accuracy is maintained despite using odometry for continuous localization.
Solution Approach 2:
The system performs preliminary sensor data collection and feature extraction during the mapping phase, storing rich environmental descriptions before navigation begins. This preliminary action allows the system to have a more accurate initial topological map that can be refined during operation, rather than relying solely on continuous odometry integration which accumulates errors.
2Measurement precision
If the topological map is updated frequently to improve accuracy, then localization precision improves, but computational complexity and processing time increase
Solution Approach 1:
The system performs full map updates only when necessary (e.g., when loop closures are detected or at scheduled intervals), rather than continuously. Between these update events, the system uses the existing topological map with odometry-based localization. This partial updating approach maintains acceptable localization precision while avoiding the continuous computational burden of full sensor-based map refinement.
Solution Approach 2:
The mapping and localization process is segmented into distinct phases: initial mapping phase where comprehensive sensor data is collected, and navigation phase where the pre-built map is used with periodic updates. This segmentation allows computationally intensive processing to be concentrated in specific phases rather than continuously, reducing overall processing complexity while maintaining localization accuracy.
3Measurement precision
If the robot explores all possible paths to verify traversability, then the topological map becomes more accurate, but the exploration time and energy consumption increase
Solution Approach 1:
The system performs preliminary traversability assessment during the initial mapping phase by attempting to navigate along potential edges and detecting obstacles with sensors. This preliminary verification allows the system to mark clearly non-traversable edges early, avoiding the need to re-explore these paths later. The robot learns from initial exploration attempts and uses this knowledge to plan more efficient subsequent explorations.
Solution Approach 2:
The system changes exploration parameters dynamically: when a new area is discovered, thorough exploration with frequent sensor checks is performed; when familiar areas are revisited, less intensive verification is used. The system adjusts the depth of traversability verification based on confidence levels from previous explorations and the importance of the path for navigation, optimizing the balance between accuracy and time consumption.
Data Source
AI summary
The operations of a computer-implemented method include obtaining a topological map of an environment including a series of waypoints and a series of edges. Each edge topologically connects a corresponding pair of adjacent waypoints. The edges represent traversable routes for a robot. The operations include determining, using the topological map and sensor data captured by the robot, one or more candidate alternate edges. Each candidate alternate edge potentially connects a corresponding pair of waypoints that are not connected by one of the edges. For each respective candidate alternate edge, the operations include determining, using the sensor data, whether the robot can traverse the respective candidate alternate edge without colliding with an obstacle and, when the robot can traverse the respective candidate alternate edge, confirming the respective candidate alternate edge as a respective alternate edge. The operations include updating, using nonlinear optimization and the confirmed alternate edges, the topological map.