LiDAR-Assisted Dead Reckoning for Robot Heading Drift Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current robotic systems face challenges in accurately determining heading angles for navigation, especially in large environments, due to the susceptibility of gyroscopes to drift and the computational complexity of methods using range sensors, which can lead to significant errors over time.
Innovation Solution
The use of LiDAR-assisted dead reckoning methods, where a robot calculates its heading angle based on velocity, time, and the orientation of objects in its environment, using a histogram to determine the primary orientation and adjust for gyroscope drift, allowing for accurate localization and mapping without relying on drift-prone instruments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If gyroscopes are used to determine heading angles for robotic navigation, then the robot can maintain orientation during movement, but the gyroscope is susceptible to drift which causes significant errors over time
Solution Approach 1:
The patent introduces LiDAR-based environmental feature detection as an intermediary system to correct gyroscope drift. The LiDAR detects surfaces and objects in the environment, and the histogram analysis of these features provides a reference frame that mediates between the drifting gyroscope data and the actual environmental orientation, thereby correcting heading angle errors without directly modifying the gyroscope itself
Solution Approach 2:
The system implements feedback by continuously comparing the gyroscope-derived heading angles with the histogram-based environmental feature orientations. When discrepancies are detected (indicating drift), the system feeds back correction signals to adjust the heading angle calculations, creating a closed-loop control system that maintains accuracy over time
2Measurement precision
If range sensors are used to calculate heading angles through scan matching algorithms, then localization accuracy can be improved, but the computational complexity increases significantly
Solution Approach 1:
The patent extracts only the essential orientation information from LiDAR scans by identifying dominant environmental features (walls, surfaces, objects) and their angular relationships. Instead of processing complete scan matching algorithms, the system extracts key geometric features and feeds them into a histogram analysis, significantly reducing computational complexity while maintaining localization accuracy
Solution Approach 2:
The system segments the environmental data into discrete angular bins for histogram analysis, dividing the continuous 360-degree space into manageable segments. This segmentation allows the robot to determine dominant environmental orientations through simple bin counting rather than complex mathematical computations, reducing processing requirements while preserving essential spatial information
3Measurement precision
If complex scan matching algorithms are used to determine robot position and orientation, then localization accuracy improves, but the processing time and computational load increase
Solution Approach 1:
The patent applies partial action by using only the necessary subset of sensor data - specifically, the angular orientations of dominant environmental features detected by LiDAR. Instead of processing complete point cloud data or performing exhaustive scan matching, the system extracts and processes only the critical orientation information needed for heading angle determination, reducing processing time while maintaining sufficient accuracy for navigation
Data Source
AI summary
Systems and methods for robotic control using LiDAR assisted dead reckoning are disclosed herein. According to at least one non-limiting exemplary embodiment, a robot may accurately localize itself over time by detecting parallelized environmental surfaces, such as walls or shelves arranged in a parallel manner, extracting a primary orientation of those surfaces using LiDAR data, and utilizing the primary orientation to accurately define its heading angle in real time, thereby enabling localization via dead reckoning.