Probabilistic Occupancy Grid Control for Mobile Industrial Robots
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for updating probability maps for mobile industrial robots in dynamic environments are inefficient, particularly when dealing with complex obstacles and varying sensor data.
Innovation Solution
A method for controlling industrial robots using a probabilistic occupancy grid, where occupancy-related measurements from elevated sensors are processed using dual models to determine new occupancy probabilities, enabling efficient data processing with RGB-D and millimeter-wave radar measurements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If sensor fusion is used to improve obstacle detection accuracy, then the robot's ability to detect difficult obstacles is improved, but the data processing complexity and computational load increase
Solution Approach 1:
The patent segments the environment into a grid of cells, with each cell having an independent occupancy probability. This segmentation allows complex sensor data to be processed cell-by-cell rather than as a whole environment, reducing computational complexity while maintaining detection accuracy. Each sensor measurement updates only the relevant cells in its field of view, not the entire grid.
Solution Approach 2:
The patent transforms raw sensor data into occupancy probability parameters for grid cells. By changing the parameter representation from raw sensor readings to probabilistic occupancy values, the system simplifies subsequent processing and decision-making. The dual-model approach further changes parameters by selecting different probability update rules based on whether objects are detected.
2Productivity
If a single probability update model is used for all measurements, then the processing is simple, but the accuracy is reduced when dealing with different types of measurements
Solution Approach 1:
The patent implements a dynamic model selection mechanism where the probability update rule changes based on the detection outcome. When an object is detected in a cell, the first model (increasing occupancy probability) is applied; when no object is detected, the second model (decreasing occupancy probability) is applied. This dynamic adaptation improves accuracy without significantly impacting processing efficiency.
Solution Approach 2:
Different probability update models are applied to different grid cells based on local measurement outcomes. Each cell receives the appropriate model based on whether its specific measurement indicated object presence or absence, rather than applying a uniform model to the entire grid. This localized approach optimizes accuracy for each measurement context.
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 allows for accurate and efficient maintenance and refinement of the occupancy grid, improving the robot's navigation and collision avoidance capabilities in complex environments.
Implementation Method 1
obtaining occupancy-related measurements using at least one measuring principle that includes sensing an incident electromagnetic wave at an elevated point on the robot
Data Source
Figure 1~3
Figure 2
Figure 4~5
AI summary
A robot (210) movable on a substrate (220) is controlled on the basis of an occupancy grid (230) of cells (231), where each cell is associated with an occupancy probability that some physical object (240, 241, 242) is present in the cell. Occupancy-related measurements are obtained by RGB-D, radar or other sensing of an incident electromagnetic wave at an elevated point (211) on the robot. From a measurement taken at an angle of incidence (S), an occupancy probability is assigned as follows: it is evaluated whether the measurement indicates that some physical object is present in the angle of incidence; a first predetermined model M1 is selected if the evaluation is positive, and second predetermined model M2 if the evaluation is negative; and a new occupancy probability for cells in the angle of incidence is determined on the basis of the measurement and the selected model.