Pedestrian Dead Reckoning Algorithm for Mobile Device Positioning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing indoor positioning systems for smartphones lack generality and robustness, often requiring context-specific tuning and failing to accurately track user movements when deviating from predefined phone placement or environmental conditions, due to assumptions about user, device, and environmental variability.
Innovation Solution
A robust pedestrian dead reckoning (PDR) algorithm using a simple phone placement classification scheme that covers all possible attachment methods, leveraging general physics principles for human motion, and incorporating a feedback loop for lifelong learning to adapt to user and environmental variations, with modules for motion classification, orientation tracking, step detection, and heading estimation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If context-specific fine-tuning is applied to optimize tracking accuracy for specific phone placements and environments, then measurement precision is improved, but device complexity and ease of operation deteriorate due to the need for extensive parameter tuning and classification
Solution Approach 1:
The patent implements a universal inertial tracking system that automatically adapts to different phone placements (handheld, pocket, wrist, etc.) and environments without requiring context-specific configuration. The system uses a classifier to automatically detect the current context and adjust parameters, making the system universally applicable across diverse scenarios while maintaining high tracking accuracy.
Solution Approach 2:
The system performs self-calibration and self-adjustment by automatically detecting user context (phone placement, walking pattern, etc.) and tuning its parameters accordingly. This eliminates the need for manual fine-tuning by users or developers for different contexts, as the system serves itself by adapting to its operating conditions in real-time.
2Measurement precision
If context-specific fine-tuning is applied to optimize tracking accuracy, then measurement precision is improved, but ease of operation worsens due to the vast tuning effort required
Solution Approach 1:
The system automatically detects the current usage context (phone placement, user behavior, environment) and adjusts its tracking parameters without requiring manual intervention. This self-service capability eliminates the vast tuning effort previously needed, making the system easy to deploy and operate while maintaining high accuracy across different contexts.
Solution Approach 2:
The system dynamically adapts its parameters based on real-time detection of user context and environmental conditions. Rather than requiring static fine-tuning for each context, the system continuously adjusts its behavior to match current conditions, making it easy to operate across diverse scenarios without manual reconfiguration.
3Measurement precision
If the system assumes knowledge of user variability parameters (height, step frequency, acceleration variance), then measurement precision is improved, but device complexity increases due to the number of parameters that must be modeled and tuned
Solution Approach 1:
The system automatically estimates and adapts user-specific parameters (height, step frequency, acceleration characteristics) during normal operation without requiring manual input or extensive pre-tuning. The classifier detects user behavior patterns and adjusts parameters on-the-fly, eliminating the complexity of manual parameter modeling while maintaining high tracking accuracy.
Solution Approach 2:
The system performs preliminary adaptation by automatically learning user characteristics during an initial period of use or during transitional phases, so that when actual tracking is needed, the parameters are already optimized for that user. This preliminary action eliminates the need for complex manual parameter setup while maintaining precision.
4Measurement precision
If the system is designed for specific phone placements (e.g., texting mode), then measurement precision is improved for that context, but adaptability deteriorates when the user deviates from the predefined placement
Solution Approach 1:
The patent creates a universal tracking system that maintains high accuracy across multiple phone placements (handheld, pocket, wrist, etc.) and environments. The classifier automatically detects the current placement and context, then adjusts parameters accordingly, making the system equally effective whether the phone is held in hand, placed in a pocket, or worn on the wrist, without requiring separate optimization for each scenario.
Solution Approach 2:
The system dynamically adapts its parameters and algorithms based on real-time detection of phone placement and user behavior. When the user changes from one placement to another, the system automatically adjusts to maintain optimal tracking accuracy, rather than being fixed to a single predefined configuration. This dynamic adaptation ensures both precision and versatility across diverse usage scenarios.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The solution provides accurate and reliable indoor tracking across various environments and user conditions, reducing errors and improving tracking accuracy by learning from experience and adjusting parameters, resulting in a more robust and adaptive indoor positioning system.
Implementation Method 1
acceleration values provided by an accelerometer
Implementation Method 2
angular velocity values provided by a gyrometer
Implementation Method 3
magnetic field values provided by a magnetometer
Data Source
Figure 1
Figure 2
Figure 3(a)~3(b)
AI summary
A mobile device has sensors providing acceleration and orientation information. These are processed to provide trajectory for the device. The processing includes improved methods of classifying the motion and detecting a user's steps as they carry the device, tracking the orientation and acceleration of the device, estimating the length of the user's steps and estimating the heading of the user. The trajectories are compared to a map and corrected by the constraints provided by the map. Parameters of the methods of producing the trajectories are optimised on the basis of maximising probability of the trajectory found by the map matching given the indicated trajectory.