Robot Route Planning for Crowded Pedestrian Avoidance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing route determination methods for autonomous mobile robots often result in frequent halts and low productivity due to unpredictable movements of pedestrians, especially in crowded environments, as they struggle to accurately avoid interference with multiple moving objects following different trajectories.
Innovation Solution
A route determination method that generates learning data by associating environmental images with action parameters, using a learned model to determine a target movement route for the robot, allowing it to navigate smoothly while avoiding interference with multiple moving objects by recognizing and predicting their movements 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 and productivity decreases
Solution Approach 1:
The system performs preliminary actions by predicting future positions of traffic participants and pre-calculating safe routes before the robot needs to halt. The route determination unit calculates multiple candidate routes in advance based on predicted interference likelihood times and virtual obstacle regions, so when an interference is predicted, the robot can immediately switch to a pre-calculated safe route without halting.
Solution Approach 2:
The system dynamically adjusts the robot's route based on real-time predictions of traffic participant movements. The route determination unit continuously monitors predicted interference likelihood times and virtual obstacle regions, and dynamically selects from multiple candidate routes the one that avoids future interference while maintaining productivity by preventing halts.
2Measurement precision
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 with unpredictable pedestrian trajectories, the robot halts for extended periods
Solution Approach 1:
The system performs preliminary route calculations by pre-determining multiple candidate routes using the A* search algorithm based on the obstacle map and grid square obstacle probabilities. These candidate routes are calculated in advance with different characteristics (e.g., different lengths, different obstacle probabilities), so when the robot needs to avoid an obstacle, it can immediately switch to a suitable pre-calculated route without extending halt time.
Solution Approach 2:
The system changes parameters by considering multiple candidate routes with different characteristics instead of a single optimal route. The route determination unit selects from candidate routes based on varying parameters such as route length, obstacle probability, and suitability for different situations, allowing the robot to maintain movement by switching routes rather than halting.
Data Source
AI summary
A method for determining a route of a robot is provided such that a moving apparatus can move smoothly to a destination point while avoiding interference with a plurality of moving objects such as traffic participants. 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 a compound environmental image constituted of time series of environmental images indicating a visual environment around a virtual robot when the virtual robot moves along each of the plurality of movement routes and a moving direction command indicating a moving direction of the virtual robot are combined, is generated as learning data. Model parameters of a CNN (action model) is learned using the learning data, and a moving velocity command for a robot is determined using a learned CNN.