Robot Localization Using Co-Located Markers and Visibility Constraints
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In warehouse environments with Automated Guided Vehicles (AGVs), accurate localization of robotic devices is challenging due to the use of markers with limited fields of visibility, which can lead to incorrect association of detected markers with their corresponding positions in the map, resulting in poor robot localization estimates.
Innovation Solution
The method involves identifying co-located markers within a threshold distance and enforcing a visibility constraint based on their non-overlapping visibility regions to accurately associate detected markers with their correct positions, improving the accuracy of robot localization by ensuring that only visible markers are considered during the data association process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If markers with limited fields of visibility are used for robot localization, then the number of markers can be reduced and device complexity is lowered, but measurement precision of robot localization deteriorates due to incorrect marker association
Solution Approach 1:
The system pre-establishes a map of markers with their positions and visibility regions before robot operation. During localization, the robot uses this pre-established information to correctly associate detected markers with map markers, preventing incorrect associations that would occur with simple proximity-based matching. This preliminary preparation enables accurate localization even with limited-visibility markers.
Solution Approach 2:
The visibility region constraint acts as an intermediary criterion between the robot's sensor detections and the map database. Instead of directly matching detected markers to nearest map markers, the system uses visibility region information as a filtering mediator to determine valid associations, ensuring that only markers within the robot's actual field of view are considered for association.
2Area of stationary object
If co-located markers are placed close together to increase localization coverage, then area coverage is improved, but reliability of marker association deteriorates due to difficulty in distinguishing between nearby markers
Solution Approach 1:
The system assigns different visibility region properties to different markers even when they are co-located at the same position. Each marker has a specific visibility region direction or orientation, allowing the robot to distinguish between them based on which marker's visibility region contains the robot's current position. This local differentiation enables reliable association even for markers at identical locations.
Solution Approach 2:
The system transitions from two-dimensional marker position matching to three-dimensional spatial reasoning by incorporating visibility region orientation. Instead of only considering planar distance between robot and markers, the system adds the vertical/directional dimension of visibility regions, allowing differentiation of co-located markers through their orientation-specific visibility cones or sectors.
Data Source
AI summary
One method disclosed includes identifying, in a map of markers fixed in an environment, two co-located markers within a threshold distance of each other, where each of the two co-located markers has a non-overlapping visibility region. The method further includes determining a set of detected markers based on sensor data from a robotic device. The method additionally includes identifying, from the set of detected markers, a detected marker proximate to a first marker of the two co-located markers. The method also includes enforcing a visibility constraint based on the non-overlapping visibility region of each of the two co-located markers to determine an association between the detected marker and a second marker of the two co-located markers. The method further includes determining a location of the robotic device in the environment relative to the map based on the determined association.