Activity Parameter Estimation via Context-Aware Sensor Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing electronic devices struggle to optimally track all aspects of an activity due to fixed parameter-determining approaches and the need for dedicated devices, lacking flexibility in selecting sensor modalities based on context and parameter-specific requirements.
Innovation Solution
A context-dependent technique for dynamically selecting sensor modalities, such as smartphones and smartwatches, to estimate activity parameters like running cadence and vertical oscillation, using a combination of device availability, context, and model selection to enhance estimation accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple dedicated devices are used to track all activity parameters, then measurement precision is improved, but device complexity and user burden increase
Solution Approach 1:
The patent enables a single electronic device to perform multiple activity parameter estimation functions by dynamically selecting from multiple sensor modalities (accelerometer, gyroscope, magnetometer, barometer, GPS, camera, microphone) and multiple estimation models. This multi-functional approach allows one device to replace what would traditionally require multiple dedicated devices, achieving comprehensive activity tracking without increasing device count.
Solution Approach 2:
The system dynamically selects which sensor modalities and estimation models to use based on current activity context, available sensors, and parameter-specific requirements. This dynamic adaptation allows the device to optimize its sensor usage in real-time, improving measurement precision for each specific parameter while avoiding the need to continuously activate all sensors, thereby managing device complexity efficiently.
2Measurement precision
If sensor modalities are dynamically selected based on context and parameter requirements, then measurement precision is improved, but computational complexity increases
Solution Approach 1:
The system performs preliminary classification of the current activity context (e.g., running, walking, cycling) and pre-identifies which sensor modalities and estimation models are most suitable for each parameter type. By pre organizing the available sensors and models into context-specific configurations, the system reduces the real-time computational burden while maintaining high estimation accuracy through context-appropriate model selection.
Solution Approach 2:
The system changes operational parameters (which sensors are active, which models are applied) based on the detected activity context and the specific parameter being estimated. This parameter adaptation allows the system to use simpler models for less critical parameters and reserve more computationally intensive models for parameters requiring higher precision, thereby balancing accuracy with computational complexity.
3Measurement precision
If dedicated devices are used for specific parameters, then measurement precision is improved, but adaptability decreases
Solution Approach 1:
The patent transforms dedicated parameter-tracking functionality into a universal capability available across all activity types. By implementing a framework that supports multiple sensor modalities and estimation models within a single device, the system can adapt to different activities (running, walking, cycling, swimming) and estimate various parameters (cadence, speed, distance, elevation) using the most appropriate sensors and models for each context, thereby achieving both precision and adaptability.
Solution Approach 2:
The system dynamically adapts its sensor and model selection based on the detected activity type and the specific parameters being tracked. This dynamic flexibility allows the same device to optimize its configuration for different activities and parameters, replacing the need for dedicated devices while maintaining high measurement precision through context-appropriate sensor and model selection.
Data Source
AI summary
In one embodiment, a method includes determining a set of one or more step-based activity parameters to estimate for a current activity of a user; and determining current parameter-estimation inputs including (1) a current activity context and (2) a current status of each of a number of sensor modalities. The method further includes selecting, for each of the one or more step-based activity parameters and based on the current parameter-estimation inputs, a respective set of one or more sensor modalities from the number of sensor modalities to use to estimate that step-based activity parameter of the current activity of the user; and for each of the one or more step-based activity parameters, estimating that activity parameter using current data from the respective selected set of one or more sensor modalities.