Robot Target Cell Selection for Adaptive Object Approach
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies struggle to accurately move a robot to a target object, especially when the target object's position changes, as they rely on predefined points and fail to adapt in real-time to new positions or obstacles.
Innovation Solution
An electronic device segments a candidate area into cells based on a robot's footprint and determines a target cell using cost, distance, and obstacle information from semantic and cost maps, allowing real-time adjustments during movement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a robot uses predefined target points for movement, then the control system is simple, but the robot cannot adapt to real-time position changes of the target object
Solution Approach 1:
The system continuously obtains updated positions of the target object through sensors and feedback loops, comparing them with the robot's current position to dynamically adjust movement commands. This allows the robot to adapt to position changes while maintaining a relatively simple control architecture through standardized feedback mechanisms.
Solution Approach 2:
The target point is transformed from a static predefined coordinate to a dynamic position that updates in real-time based on the target object's location. The system calculates the robot's pose and dynamically determines the next target position, enabling adaptability without requiring complete reprogramming of the control system.
2Measurement precision
If a robot calculates final target position based on object position, then movement accuracy improves, but real-time reflection of position changes becomes challenging
Solution Approach 1:
The system pre-calculates the relationship between target object position and robot target points, storing transformation rules and cost map data in advance. When the target object moves, the system only needs to apply pre-established calculation methods to new position data, reducing real-time computational burden and maintaining accuracy.
Solution Approach 2:
The system maintains continuous operation by updating the target position incrementally as the robot moves and the target object relocates. Rather than performing complete recalculation cycles, the system continuously adjusts the target point based on new sensor data, ensuring both accuracy and real-time responsiveness.
3Measurement precision
If a robot divides the area into multiple cells, then navigation precision improves, but the computational complexity increases
Solution Approach 1:
The environment is divided into a grid of cells with the robot's footprint applied to determine valid positioning cells. This segmentation allows precise navigation by evaluating costs for each cell, while the regular grid structure maintains computational efficiency through standardized processing routines for each cell evaluation.
Solution Approach 2:
The system assigns different cost values to different cells based on local characteristics such as proximity to the target object, obstacles, and robot orientation requirements. This allows high precision navigation by considering local conditions in each cell while maintaining overall system efficiency through localized cost evaluation rather than global optimization.
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
This approach enables precise and adaptive robot movement to a target object by identifying and avoiding obstacles, ensuring accurate navigation even when the target object's position changes.
Implementation Method 1
based on a map-based global cost map generated by use of a light detection and ranging (LiDAR) sensor
Data Source
AI summary
In the electronic device, and a control method thereof, the at least one processor can obtain a sub-map, from the semantic map, including a candidate area in which a robot is able to be located according to a movement command, where the candidate area is within a set distance relative to a position of the target object, segment the candidate area into a plurality of cells, based on a footprint of the robot, and determine a target cell among the plurality of cells, based on one of or any combination of a first input about a cost of each of the plurality of cells, a second input about a cell-target distance between each of the plurality of cells and the target object, or a third input about a cell-robot distance between each of the plurality of cells and the robot.