Upward Facing Sensor SLAM for Drift-Free Navigation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional Simultaneous Localization and Mapping (SLAM) techniques for robots, especially in GPS-degraded environments like warehouses, face significant errors due to estimation drift and reliance on loop closure, which can lead to catastrophic failures and degraded map quality, especially in long-term navigation tasks.
Innovation Solution
A method and system utilizing an upward-facing imaging sensor to compute vehicle attitude combined with range-bearing measurements from sensors like LiDAR or sonar to determine accurate position and map estimates, incorporating inertial sensors and a Kalman filter for robust localization and mapping, allowing robots to navigate without prior knowledge of their environment and without relying on loop closure.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional SLAM techniques are used for long-term navigation, then the robot can operate without GPS, but position accuracy degrades significantly over long trajectories due to estimation drift
Solution Approach 1:
The patent introduces an upward-facing imaging sensor as an intermediary device that captures images of the ceiling environment. These images serve as intermediate data that, when processed through attitude computation algorithms, provide absolute orientation information. This intermediary measurement mechanism breaks the chain of cumulative error in conventional SLAM by providing independent orientation references.
Solution Approach 2:
The patent replaces the mechanical/odometric-based orientation estimation (which accumulates drift) with an optical-based attitude computation system. By using upward-facing images and computing device attitude through image processing and geometric relationships, the system substitutes the drift-prone mechanical sensing approach with an optical measurement approach that provides absolute orientation references.
2Measurement precision
If loop closure is used to correct estimation drift, then localization accuracy can be recovered, but the system becomes sensitive to data association errors which can cause catastrophic failure
Solution Approach 1:
The patent performs preliminary attitude computation continuously during navigation using upward-facing images. By pre-computing and maintaining absolute orientation information throughout the trajectory, the system eliminates the need for later loop closure operations. This preliminary action of continuous attitude estimation prevents drift accumulation before it becomes significant, removing the requirement for risky data association operations.
Solution Approach 2:
The patent extracts and removes the loop closure component from the SLAM system by introducing independent absolute orientation measurements. By taking out the dependency on loop closure for drift correction and replacing it with continuous attitude computation from upward-facing images, the system eliminates the vulnerability to data association errors while maintaining localization accuracy.
3Measurement precision
If more sensors are added to improve localization accuracy, then position estimation improves, but device complexity increases
Solution Approach 1:
The patent makes the upward-facing imaging sensor multi-functional by using it for both navigation mapping and attitude computation. The same sensor that captures environmental features for SLAM also provides absolute orientation information when processed through the attitude computation algorithm. This universality eliminates the need for separate dedicated orientation sensors, maintaining measurement precision while controlling device complexity.
Data Source
AI summary
Described herein are embodiments of a method and system that uses a vertical or upward facing imaging sensor to compute vehicle attitude, orientation, or heading and combines the computed vehicle attitude, orientation, or heading with range bearing measurements from an imaging sensor, LiDAR, sonar, etc., to features in the vicinity of the vehicle to compute accurate position and map estimates.