Robot Visual Odometry for Autonomous Sensor Bias Calibration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current robots face challenges in maintaining accurate sensor calibration over time due to wear and tear, collisions, and environmental factors, leading to unreliable data for localization and navigation.
Innovation Solution
The implementation of laser and imaging odometry systems that allow robots to autonomously determine extrinsic biases in sensors by analyzing image discrepancies between images captured at different times, thereby enhancing localization capabilities and reducing reliance on classical odometry units.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If classical odometry units (gyroscopes, accelerometers) are used to measure robot motion, then motion data can be obtained, but the data becomes unreliable over time due to biases and noise
Solution Approach 1:
The system continuously compares odometry-based motion estimates with visual flow measurements from the sensor, using the discrepancy as feedback to correct odometry biases. This closed-loop feedback mechanism compensates for drift and maintains long-term reliability of motion data.
Solution Approach 2:
The patent replaces reliance on mechanical odometry units (gyroscopes, accelerometers) with an optical-based visual odometry system that uses image sequences to determine sensor pose changes. This substitution eliminates mechanical biases and noise inherent in traditional odometry units.
2Measurement precision
If operator verification of sensor calibration is performed, then calibration accuracy can be maintained, but the process becomes time consuming and impractical
Solution Approach 1:
The robot performs self-calibration by autonomously analyzing visual flow from sensor images and comparing it with odometry data. The system automatically detects and corrects extrinsic parameter biases without human intervention, making the calibration process independent and continuous rather than periodic and manual.
Solution Approach 2:
The system continuously performs calibration verification in the background during normal operation, rather than waiting for scheduled manual checks. This preliminary and ongoing calibration ensures accuracy is maintained proactively rather than reactively.
3Productivity
If sensor calibration is not verified, then operation continues without interruption, but calibration degradation leads to inaccurate localization and navigation
Solution Approach 1:
The system continuously performs calibration verification and correction during normal robot operation, ensuring that localization accuracy is maintained without interrupting productivity. The calibration process runs concurrently with navigation tasks rather than requiring separate calibration sessions.
4Reliability
If multiple sensors and odometry units are used, then motion measurement redundancy is achieved, but system complexity increases
Solution Approach 1:
The patent merges data from multiple sensors (camera, depth sensor, odometry units) into a unified visual odometry framework. By combining these sources and using their discrepancies for calibration, the system achieves reliable motion measurement while managing complexity through integrated processing rather than separate independent systems.
Data Source
AI summary
Systems and methods for laser and imaging odometry for autonomous robots are disclosed herein. According to at least one non-limiting exemplary embodiment, a robot may utilize images captured by a sensor and encoded with a depth parameter to determine its motion and localize itself The determined motion and localization may then be utilized to verify calibration of the sensor based on a comparison between motion and localization data based on the images and motion and localization data based on data from other sensors and odometry units of the robot.