Trajectory-based localization and mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current localization and mapping techniques for autonomous robots, such as cleaning robots, face challenges in accurately re-localizing themselves in environments with complex features, often relying on straight wall-based methods that limit the number of possible landmarks and require complex navigation routines, while also increasing computational burdens and costs.
Innovation Solution
The method involves generating and storing trajectory landmarks that characterize curved or non-straight paths around environmental features, using sensors like encoders, bump sensors, and gyroscopes to create trajectory-based re-localization, which allows for efficient re-localization by matching new trajectories with pre-stored landmarks, correcting pose estimates, and updating internal maps.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If straight wall-based localization methods are used, then the localization process is simplified, but the number of possible landmarks is limited and navigation complexity increases
Solution Approach 1:
The patent applies curvature by using curved or non-straight trajectories as landmarks instead of straight walls. The robot follows curved paths around environmental features and uses these curved trajectories as reusable landmarks for re-localization, thereby increasing the number of possible landmarks without increasing navigation complexity
Solution Approach 2:
The patent makes trajectory data serve multiple functions: it is used both for navigation (path planning) and for re-localization (landmark matching). By storing trajectory data in a reusable format, the same data structure serves dual purposes, reducing overall system complexity while increasing versatility
2Measurement precision
If complex localization algorithms are used to handle curved trajectories, then re-localization accuracy improves, but computational burden increases
Solution Approach 1:
The patent performs preliminary action by pre-processing and storing trajectory data in a standardized, reusable format during the initial navigation phase. This pre-processing includes capturing sensor data and organizing it into a format suitable for later matching, which reduces the computational burden during actual re-localization operations while maintaining high accuracy
Solution Approach 2:
The patent uses copying by creating reusable copies of trajectory data that can be stored and referenced multiple times. Instead of re-computing localization from scratch each time, the system copies and matches against previously stored trajectory landmarks, significantly reducing computational burden while maintaining re-localization accuracy
3Measurement precision
If more sensors are added to improve localization accuracy, then re-localization precision improves, but device cost and complexity increase
Solution Approach 1:
The patent applies self-service by using the robot's existing navigation sensors (encoders, bump sensors, gyroscopes) to simultaneously perform both navigation and localization functions. The system serves itself by reusing already-collected sensor data from normal operation, eliminating the need for additional dedicated localization sensors and reducing overall device complexity
4Loss of time
If traditional SLAM methods are used, then mapping is achieved, but re-localization time increases in complex environments
Solution Approach 1:
The patent performs preliminary action by pre-processing trajectory data during initial navigation and storing it in an optimized format for rapid matching. This advance preparation includes organizing sensor data into reusable landmark representations, which significantly reduces re-localization time when the robot needs to re-localize in complex environments without compromising mapping efficiency
Data Source
AI summary
An autonomous robot is maneuvered around a feature in an environment along a first trajectory. Data characterizing the first trajectory is stored as a trajectory landmark. The autonomous cleaning robot is maneuvered along a second trajectory. Data characterizing the second trajectory is compared to the trajectory landmark. Based on comparing the data characterizing the second trajectory to the trajectory landmark, it is determined that the first trajectory matches the second trajectory. A transform that aligns the first trajectory with the second trajectory is determined. The transform is applied to an estimate of a position of the autonomous cleaning robot as a correction of the estimate.