Neural Network Local Motion Planning for Real-Time Collision Avoidance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for generating local motion for mobile objects, such as robots, in complex environments lack efficient collision avoidance and path planning, particularly in dynamic settings where real-time obstacle information is crucial.
Innovation Solution
A method and apparatus that utilize a neural network to determine a local window for a mobile object, generate local cost and goal maps, and calculate a target velocity based on input data including these maps and the object's current velocity, integrating sensor data for real-time obstacle information, to navigate through the environment while minimizing collision probability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional path planning methods are used for mobile objects in complex environments, then the system structure is simple, but the collision avoidance capability and path planning efficiency are insufficient in dynamic settings
Solution Approach 1:
The patent replaces traditional mechanical path planning algorithms with a neural network-based system. The neural network processes local cost maps and sensor data to generate target velocities, substituting conventional control mechanisms with machine learning-based decision making. This enables adaptive collision avoidance in dynamic environments while maintaining system architecture.
Solution Approach 2:
The patent segments the environment into local windows around the mobile object, creating local cost maps for each segment. This segmentation allows the system to process complex environments in manageable portions, improving real-time collision avoidance performance without overwhelming system complexity.
2Productivity
If real-time obstacle information is processed using traditional methods, then the processing speed is limited, but the path planning accuracy can be maintained
Solution Approach 1:
The patent performs preliminary actions by pre-processing sensor data into local cost maps and pre-training the neural network on simulated environments. This preparation enables the system to quickly process real-time obstacle information during operation, achieving both high processing speed and accurate path planning without compromising either parameter.
3Reliability
If a neural network is introduced for local motion generation, then the collision avoidance and path planning capabilities are improved, but the computational complexity increases
Solution Approach 1:
The patent applies local quality by focusing the neural network's computational effort on local windows around the mobile object rather than processing the entire environment. The local cost maps provide localized information that the neural network processes to generate target velocities, achieving high navigation safety while reducing overall computational complexity through spatial localization.
Data Source
AI summary
Apparatuses and methods for generating a local motion of a mobile object based on a machine learning are disclosed. The methods may include determining, by processing circuitry, a local window corresponding to a current position of the mobile object in a global area; generating, by processing circuitry, a local cost map indicating a probability of a collision of the mobile object with an obstacle in the local window; generating, by processing circuitry, a local goal map indicating a local path in the local window between the current position and a local goal in the local window; and determining, by processing circuitry, a target velocity of the mobile object based on output data of a neural network, the output data based on an input of input data to the neural network, the input data including the local cost map, the local goal map, and a current velocity of the mobile object.