Robot Route Learning for Crowded Pedestrian Navigation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing route determination methods for autonomous mobile robots often result in frequent interference and halts due to unpredictable movements of pedestrians, particularly in crowded environments, leading to low productivity.
Innovation Solution
A route determination method that generates learning data by recognizing movement routes while avoiding interference, using environmental images and action parameters to create a learned model for determining target movement routes, allowing the robot to move smoothly to a destination while avoiding multiple moving objects with different movement patterns.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If predicted interference likelihood times and virtual obstacle regions are used to determine robot routes, then interference with traffic participants is avoided, but the robot halts frequently when pedestrians move along unpredictable trajectories, causing low productivity
Solution Approach 1:
The system performs preliminary actions by calculating predicted interference likelihood times and virtual obstacle regions in advance based on current pedestrian velocities and positions. This allows the robot to proactively plan routes that avoid future interference points rather than reacting to actual collisions, reducing unnecessary halts while maintaining safety
Solution Approach 2:
The route determination system dynamically updates predicted interference likelihood times and virtual obstacle regions as pedestrian velocities and positions change. This dynamic recalculation allows the robot to adapt to unpredictable pedestrian movements in real-time, maintaining reliable interference avoidance while minimizing halts by continuously optimizing the route based on current conditions
2Ease of operation
If grid squares with lowest obstacle probability are determined using A* search algorithm, then routes are calculated based on obstacle maps, but in crowded environments where unpredictable pedestrian movements occur, grid squares cannot be found and robot halt time is lengthened
Solution Approach 1:
The system calculates predicted interference likelihood times and virtual obstacle regions in advance for multiple future time points, creating a forward-looking obstacle avoidance framework. This preliminary calculation allows the robot to identify safe passage times before reaching potential conflict zones, eliminating the need for time-consuming A* search recalculations when pedestrians move unpredictably
Solution Approach 2:
The invention replaces the traditional A* search algorithm on static grid maps with a dynamic prediction-based approach. Instead of repeatedly searching for paths on updated obstacle maps, the system substitutes this with velocity-based predictions of pedestrian trajectories and interference likelihood calculations, dramatically reducing computation time and robot halts in crowded environments
Data Source
AI summary
In an environment in which a plurality of second pedestrians moves along predetermined movement patterns, a plurality of movement routes when a first pedestrian moves toward a destination point is recognized. Data, in which an environmental image indicating a visual environment in front of a virtual robot when the virtual robot moves along each of the movement routes and a moving direction command indicating a moving direction of the virtual robot are combined, is generated as learning data. In the environmental image, colors corresponding to kinds of the moving objects are applied to at least a portion of moving object image regions indicating pedestrians (moving objects) present around a robot. Model parameters of a CNN (action model) is learned using the learning data, and a moving velocity command for the robot is determined using a learned CNN.