Mobile Agent SLAM Mapping With Manhattan Orientation Constraints
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional SLAM algorithms for determining the location of mobile agents suffer from orientation drift due to measurement errors, particularly in environments with regular structures like buildings, where compass measurements are often unavailable, leading to inaccuracies in environment mapping.
Innovation Solution
Combining the SLAM algorithm with Manhattan world assumptions, incorporating Manhattan nodes and edges into the SLAM graph, and using optimization methods to minimize error functions based on Manhattan orientations, allowing for precise alignment of the reference coordinate system with the environment's structure.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a conventional SLAM algorithm is used to determine the pose of the mobile agent, then the environment map can be created, but orientation drift accumulates due to measurement errors
Solution Approach 1:
The patent changes the parameter representation by introducing Manhattan orientation (discrete angular parameters) alongside continuous pose parameters. This discretization of orientation into Manhattan-aligned angles reduces the accumulation of drift errors by constraining the solution space to physically meaningful orientations in structured environments.
Solution Approach 2:
The patent introduces Manhattan orientation as an intermediary parameter that mediates between sensor measurements and the final pose estimation. This intermediary helps to bridge the gap by providing a reference frame that is consistent with the environment structure, thereby reducing orientation drift.
2Reliability
If direct compass measurements are used to compensate for orientation drift, then orientation accuracy improves, but the system becomes unavailable in environments with stray magnetic fields or without compass sensors
Solution Approach 1:
The system uses the environment structure itself (walls and objects with regular geometries) to provide orientation reference information. Instead of relying on external compass measurements, the system extracts Manhattan orientation cues from the captured sensor data of the environment, making the system self-sufficient and adaptable to various environments.
Solution Approach 2:
The patent replaces the magnetic field-based compass measurement system with a vision/LIDAR-based system that infers orientation from environmental structures. This substitution eliminates dependency on magnetic fields and compass hardware, improving adaptability to environments where compass measurements are unavailable or unreliable.
3Area of stationary object
If the SLAM algorithm processes all sensor data without constraints, then complete environment mapping is achieved, but computational complexity and processing time increase
Solution Approach 1:
The patent applies local quality by introducing Manhattan constraints specifically at locations where environmental structures (walls, objects) are detected, rather than uniformly across the entire environment. This localized application of constraints reduces computational overhead while maintaining mapping accuracy in structured regions.
Solution Approach 2:
The patent segments the environment processing into regions with Manhattan structures and regions without, applying different processing strategies. In Manhattan regions, constrained optimization is used to reduce computational complexity, while other regions use standard SLAM processing, thereby balancing completeness and efficiency.
Data Source
AI summary
A computer-implemented method for operating a mobile agent in an environment on the basis of a pose of the mobile agent. The location of the mobile agent is determined by way of the following steps in order to operate the mobile agent: capturing sensor data regarding walls and/or objects located in an environment, wherein the sensor data indicate the orientation and distance of the walls and/or objects in an agent-based agent coordinate system; establishing the pose of the mobile agent in the environment using a SLAM algorithm while taking account of a Manhattan orientation.