Fault Detection and Recovery for Concurrent Odometry Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Machine vision techniques for augmented and virtual reality, such as SLAM, face delays and inaccuracies due to faults in visual and inertial sensors when tracking motion, leading to disruptions in building accurate maps of the environment.
Innovation Solution
An electronic device equipped with a front-end motion tracking module and a back-end mapping module that detects fault conditions in sensor data, buffers or discards compromised data, and merges maps generated before and after the fault to maintain accurate tracking and mapping, using image and non-image sensor data to correct pose estimation and build a three-dimensional representation of the environment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If visual and inertial sensors are used for tracking motion in SLAM, then motion tracking capability is improved, but faults in sensors cause delays and inaccuracies in map building
Solution Approach 1:
The system performs preliminary actions by detecting fault conditions in sensor data before they can cause significant delays or inaccuracies in map building. The fault detection mechanism monitors sensor data quality continuously and proactively identifies compromised data, allowing the system to take preventive measures such as switching to alternative data sources or adjusting processing priorities before the fault impacts map building performance.
Solution Approach 2:
The system introduces an intermediary fault detection and recovery mechanism between the visual/inertial sensors and the map building process. This intermediary layer analyzes sensor data for faults, manages the flow of data to the mapping algorithms, and ensures that only reliable data is used, thereby protecting the map building reliability while maintaining the speed benefits of using multiple sensors.
2Productivity
If sensor data is processed continuously for real-time mapping, then map building speed is improved, but faults in sensor data cause inaccuracies
Solution Approach 1:
The system implements feedback mechanisms where the fault detection module continuously monitors sensor data quality and provides feedback to the map building process. When faults are detected, the feedback loop triggers corrective actions such as discarding compromised data, switching to backup sensors, or adjusting processing parameters, thereby maintaining map accuracy without sacrificing continuous processing capability.
Solution Approach 2:
The system dynamically changes processing parameters based on detected fault conditions. When faults are identified in sensor data, the system adjusts parameters such as data acceptance thresholds, processing priorities, or sensor selection to exclude compromised data while maintaining continuous operation. This allows the system to preserve map building speed while ensuring that only accurate data contributes to the map.
3Measurement precision
If fault detection and data buffering mechanisms are implemented, then map accuracy is improved, but system complexity increases
Solution Approach 1:
The system segments the SLAM pipeline into distinct functional modules: sensor data acquisition, fault detection, data validation, buffering/discarding logic, and map building. This segmentation allows each component to be optimized independently and facilitates easier debugging and maintenance. The fault detection and recovery mechanisms are implemented as separate modular components that can be added or removed without affecting the core SLAM algorithms, thereby managing complexity through structured organization.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An electronic device (100) detects and recovers from fault conditions while tracking its motion and building a map (417) of its environment (112). A front-end motion tracking module (210) detects fault conditions while tracking motion over time and provides mapping data (312, 316) to a back-end mapping module (220). The motion tracking module indicates to the mapping module when a fault condition has been detected and when the fault condition is no longer detected. The mapping module generates maps (417) from the mapping data and merges the maps to build a three-dimensional representation (225) of the environment. The mapping module buffers or discards any mapping data received from the motion tracking module during the existence of a fault condition. The mapping module merges the maps generated based on mapping data received before and after the fault condition and adds the merged maps to the three-dimensional representation of the environment.