IoT Location History Filtering via Outlier Removal
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing IoT tracking systems face challenges in accurately determining the location of objects, especially when GPS or Wi-Fi signals are unreliable, leading to incorrect location histories and inaccurate estimated times of arrival (ETAs).
Innovation Solution
A multi-layer outlier location filtering process is employed, which involves identifying and merging adjacent location reports indicating no movement, filtering out outliers based on cell phone tower identifiers, threshold distances, and sensor data, and removing single extreme and triangle outlier locations to generate a reliable location history.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If GPS or Wi-Fi location reporting is used to track objects, then location tracking capability is improved, but signal dropout and transmitter offline issues reduce measurement precision
Solution Approach 1:
The patent combines multiple location determination methods (GPS, Wi-Fi, cell tower triangulation, inertial sensors) into a unified location tracking system. When GPS signals drop out, the system merges alternative location sources to maintain continuous tracking, thereby improving reliability without sacrificing measurement precision.
Solution Approach 2:
The system performs preliminary filtering and validation of location data before it is used for tracking. By pre-processing location reports to identify and remove outliers, the system ensures high measurement precision even when signal conditions are poor or transmitters go offline.
2Measurement precision
If multiple location reports are collected to improve tracking accuracy, then measurement precision is improved, but data complexity and processing requirements increase
Solution Approach 1:
The patent segments the location data processing into distinct stages: initial outlier filtering based on geometric criteria (triangle outliers, vertex outliers), intermediate filtering using motion constraints, and final validation. This segmentation reduces overall data processing complexity by handling different aspects of data quality separately.
Solution Approach 2:
The system changes parameters of location reports dynamically based on signal quality and motion state. By adjusting weighting factors, threshold values, and validation criteria according to current conditions, the system maintains high measurement precision while adapting processing complexity to actual needs.
3Measurement precision
If location data is filtered to remove outliers, then measurement precision is improved, but loss of potentially valid location information occurs
Solution Approach 1:
The patent implements feedback mechanisms where filtered location data is continuously validated against expected motion patterns and physical constraints. The system uses feedback from multiple sensors (accelerometers, gyroscopes) to determine whether apparent outliers are truly erroneous or represent legitimate rapid movements, thereby minimizing information loss while maintaining precision.
Solution Approach 2:
The filtering criteria are dynamic rather than static. The system adapts outlier thresholds and validation rules based on current motion state, signal quality, and environmental conditions. This dynamic approach ensures that valid location information is preserved during legitimate rapid movements while still filtering actual errors, reducing information loss.
Data Source
AI summary
A method for generating a location history for a tracked object includes merging two or more location reports that are adjacent in time and associated with indications that the tracked object has not moved; discarding one of two or more location reports that are adjacent in time and associated with different cell phone tower identifiers; discarding a location report that is greater than a first threshold distance from an immediately preceding location report and from an immediately following location report; and discarding a location report that is greater than a second threshold distance from an immediately preceding location report when both reports are associated with indications that the tracked object has not moved.