Robot Occupancy Mapping by Filtering Dynamic and Unfixed Objects
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Mobile robots face challenges in constructing accurate occupancy maps due to objects that adversely affect map accuracy, such as dynamic and unfixed objects which are difficult to distinguish from static and fixed objects using conventional sensors.
Innovation Solution
A robot system that classifies objects into predefined permanency categories (fixed, static and unfixed, dynamic) using image analysis, excluding undesirable classifications to generate a more accurate occupancy map, with the ability to update classifications based on repeated observations and user-defined criteria.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional sensors are used to detect objects in the region, then the robot can detect objects present in the environment, but the robot cannot reliably distinguish between static/fixed objects and dynamic/unfixed objects, resulting in inaccurate occupancy maps
Solution Approach 1:
The patent segments the occupancy map generation process into distinct phases: initial mapping phase where all detected objects are included, and operational phase where only static and fixed objects are retained. This segmentation allows the system to handle the contradiction by treating different object types differently at different stages of operation.
Solution Approach 2:
The patent implements dynamic object classification where objects are initially assumed to be part of the environment (static/fixed) and are reclassified as dynamic/unfixed when detected in multiple different locations during robot operations. This dynamic adaptation resolves the contradiction by allowing the system to maintain high measurement precision initially while ensuring reliability through continuous observation and reclassification.
2Quantity of substance
If the robot includes all detected objects in the occupancy map, then the map is comprehensive, but dynamic and unfixed objects adversely affect the accuracy and reliability of the map
Solution Approach 1:
The patent extracts and removes dynamic and unfixed objects from the occupancy map after initial construction. During the operational phase, when objects are detected in multiple locations, they are identified as dynamic/unfixed and excluded from the final occupancy map, retaining only static and fixed objects. This extraction process maintains comprehensive initial mapping while ensuring final map reliability.
Solution Approach 2:
The patent implements a feedback mechanism where the robot continuously monitors object positions during operations and uses this information to reclassify objects. Objects detected in multiple different locations provide feedback that they are dynamic/unfixed, triggering their exclusion from the occupancy map. This feedback loop ensures map accuracy while maintaining comprehensive object detection.
3Measurement precision
If the robot traverses the region multiple times to improve map accuracy, then more objects can be classified correctly, but time is lost and productivity decreases
Solution Approach 1:
The patent performs preliminary occupancy map construction during the initial robot traversal, including all detected objects in the map. This preliminary action ensures comprehensive coverage without requiring multiple traversals for basic map creation. Subsequent traversals are used only for object reclassification, significantly reducing the time penalty while maintaining high detection accuracy.
Solution Approach 2:
The patent applies partial action by focusing subsequent traversals specifically on object reclassification rather than complete remapping. Only objects detected in multiple locations are reevaluated and reclassified, while the rest of the map remains unchanged. This partial reprocessing approach maintains high measurement precision without the full time cost of complete remapping.
Data Source
AI summary
An example control system includes a memory and at least one processor to obtain image data from a given region and perform image analysis on the image data to detect a set of objects in the given region. For each object of the set, the example control system may classify each object as being one of multiple predefined classifications of object permanency, including (i) a fixed classification, (ii) a static and fixed classification, and/or (iii) a dynamic classification. The control system may generate at least a first layer of a occupancy map for the given region that depicts each detected object that is of the static and fixed classification and excluding each detected object that is either of the static and unfixed classification or of the dynamic classification.