Accelerometer Data Storage Selection via Activity Ratio
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Activity trackers face challenges in efficiently storing accelerometer data due to limited storage capacity and the need to differentiate between active and inactive periods, as distinguishing step activity from other motions is difficult.
Innovation Solution
A processor-implemented system that selects and stores accelerometer data by grouping readings into data windows, identifying local maxima, and calculating an activity ratio to determine which periods to store, based on count and ratio thresholds, thereby optimizing storage by focusing on pertinent time periods.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If all accelerometer data is stored indiscriminately, then complete activity information is preserved, but storage capacity is quickly exhausted
Solution Approach 1:
The patent segments accelerometer data into discrete time windows and applies different storage decisions to each window based on activity detection. This allows selective storage of only those windows containing relevant activity patterns, preserving important information while reducing overall storage requirements.
Solution Approach 2:
The patent applies different quality thresholds and storage priorities to different time periods based on detected activity characteristics. Windows containing activity patterns matching target behaviors (e.g., walking, running) are stored with higher priority, while inactive periods are discarded, creating local quality variations in the stored data set.
2Loss of information
If data from all time periods is stored, then comprehensive activity coverage is achieved, but storage efficiency decreases
Solution Approach 1:
The patent performs preliminary analysis of accelerometer data windows to detect activity patterns before making storage decisions. By预先 identifying windows containing relevant activity characteristics, the system can selectively store only those periods, achieving comprehensive coverage of important activities while maintaining high storage efficiency.
3Measurement precision
If activity detection thresholds are lowered to capture more step activity, then detection sensitivity increases, but false positives from other motions increase
Solution Approach 1:
The patent applies multiple overlapping detection criteria and requires activity patterns to satisfy multiple conditions simultaneously (e.g., acceleration magnitude, frequency characteristics, temporal patterns). This partial action approach with multiple criteria reduces false positives while maintaining sensitivity to genuine step activity.
Solution Approach 2:
The system uses feedback from multiple sensor readings and historical activity patterns to adjust detection decisions. By analyzing sequences of windows and comparing against learned activity patterns, the system can distinguish true step activity from other motions more reliably, improving both detection precision and classification reliability.
Data Source
AI summary
Systems and methods select accelerometer data to store on computer-readable media. Systems may comprise an accelerometer, a processor, and a computer-readable medium storing instructions executed by the processor. Accelerometer data from a critical time period are stored on a computer readable medium if an activity ratio for the critical time period exceeds an activity threshold. The numerator of the activity ratio comprises a count of data windows in the critical time period that satisfy one or more count criteria and the denominator comprises the total number of data windows in the critical time period.