Cadence Determination in Wearables via Dynamic Thresholding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methodologies for determining cadence in mobile and wearable devices are prone to errors, especially with weak or noisy motion signals, particularly at low walking velocities, which is significant in healthcare applications and leads to inaccuracies.
Innovation Solution
The use of advanced signal processing techniques such as wavelet transforms and sensor fusion to accurately determine cadence by analyzing motion sensor signals, combining time and frequency domain analysis to filter out noise and improve accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If threshold-based detection methods are used to determine cadence, then the device complexity is low and ease of operation is high, but measurement precision deteriorates especially at low walking velocities
Solution Approach 1:
The patent implements dynamic adaptation of detection thresholds based on signal characteristics. The system continuously adjusts thresholds according to the detected signal strength and noise levels, transitioning from static thresholding to dynamic thresholding. This allows the system to maintain high measurement precision across varying walking velocities while managing computational complexity through adaptive rather than purely complex algorithms.
Solution Approach 2:
The patent changes key parameters of the detection system including threshold values, analysis window sizes, and frequency bands based on signal conditions. By dynamically adjusting these parameters rather than using fixed values, the system achieves high precision across different walking speeds without requiring overly complex processing for each condition.
2Measurement precision
If simple thresholding algorithms are used, then processing speed is high and productivity is maintained, but measurement precision deteriorates with weak or noisy motion signals
Solution Approach 1:
The patent divides the signal processing into distinct segments: initial signal acquisition, quality assessment, threshold determination, and cadence detection. This segmentation allows the system to apply computationally intensive precision algorithms only when signal quality warrants it, while maintaining faster processing for clear signals, thus balancing precision with processing time efficiency.
Solution Approach 2:
The patent applies partial complexity - using simple thresholding when signals are clear and strong, but applying more sophisticated frequency-domain analysis and adaptive thresholding only when signal quality deteriorates. This partial application of complex algorithms maintains average processing time while improving precision for the critical cases of weak or noisy signals.
3Measurement precision
If frequency domain analysis is added to improve precision, then measurement precision improves, but device complexity increases
Solution Approach 1:
The patent dynamically selects between time-domain and frequency-domain analysis methods based on signal characteristics. When signals are strong and clear, simple time-domain thresholding is used. When signals are weak or noisy, the system transitions to frequency-domain analysis with FFT to extract cadence information. This dynamic method selection improves precision when needed while minimizing average computational complexity.
Data Source
AI summary
Methods for controlling in real time an aspect of a process in a mobile or wearable device with a cadence of the device user are described, where the cadence of the device user is determined with an update frequency greater than the user's step frequency, and the controlling is also performed with an update frequency greater than the user's step frequency. For example, the determination of the user's cadence may leverage a combination of frequency and time techniques, with the analysis of, for instance, substantial orientation changes experienced by the device, or the detection of abrupt changes in the user's cadence.