Transition Detection Using Trajectory Data and Map References
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Global Navigation Satellite Systems (GNSS) struggle to accurately determine transitions between indoor and outdoor zones, such as doorways, due to reduced accuracy in close proximity to buildings, leading to incorrect identification of entry and exit points.
Innovation Solution
A computer-implemented method and apparatus that process trajectory data from mobile computing devices to determine the location and time of transitions by positioning trajectories in a frame of reference relative to a map, using signal gradients and Simultaneous Localisation and Mapping techniques, to identify candidate transitions and determine accurate entry and exit points.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If GNSS data is used to determine position, then position information is obtained, but accuracy deteriorates in close proximity to buildings
Solution Approach 1:
The patent introduces map data as an intermediary reference system. Instead of relying solely on direct GNSS measurements which are blocked near buildings, the system uses map outlines and trajectory data as intermediate references to infer transition points, thereby overcoming the signal blockage problem.
Solution Approach 2:
The patent creates a virtual representation of the physical environment by overlaying map data (outlines of buildings and zones) onto the trajectory data. This copied spatial information from maps serves as a reference to accurately identify transitions without requiring direct GNSS accuracy at the transition point itself.
2Reliability
If map outlines are used without entrance locations, then building boundaries are defined, but transition points cannot be identified
Solution Approach 1:
The patent performs preliminary actions by collecting and storing trajectory data from multiple mobile devices before attempting to identify transitions. This accumulated trajectory information is then processed to infer entrance locations, effectively performing the information extraction in advance rather than requiring it to be pre-mapped.
Solution Approach 2:
The system uses feedback from multiple trajectory observations to refine the identification of transition points. By analyzing where multiple trajectories converge or cross building boundaries, the system iteratively improves the accuracy of inferred entrance locations, using the collective information as feedback to overcome the initial lack of entrance data.
3Measurement precision
If multiple trajectories are processed, then transition accuracy improves, but computational complexity increases
Solution Approach 1:
The patent segments the processing task by first filtering and preprocessing individual trajectories to identify candidate transition points, then grouping these candidates across multiple trajectories. This segmentation divides the complex overall task into manageable stages: individual trajectory analysis, candidate identification, and aggregation, reducing the computational burden at each step.
Solution Approach 2:
The patent applies partial action by focusing processing only on relevant portions of trajectory data near building boundaries rather than analyzing entire trajectories. By concentrating computational resources on the critical regions where transitions are likely to occur, the system achieves high accuracy without the excessive computational cost of processing all trajectory data in full detail.
Data Source
AI summary
A computer implemented method (400) of determining a location of one or more transitions (5a-e) on a map and/or the time at which one or more transitions (5a-e) occurs, the one or more transitions (5a-e) made by a set of mobile computing devices (13a-c), from a first zone (2) to a second zone (4), the method (400) comprising: obtaining (402) trajectory data representing a plurality of trajectories (302a, b) collected from one or more mobile computing devices (13a-c), at least some of the trajectories (302a, b) passing through the first zone (2) and/or the second zone (4); positioning (404) the trajectories (302a, b) in a frame of reference defined relative to the map, wherein at least part of at least some of the trajectories (302a, b) are positioned based on correspondence with other trajectories (302a, b); and processing the plurality of trajectories (302a, b) positioned in the frame of reference defined relative to the map to determine the location and/or time.