Accelerometer Step Detection for Walking and Running Distinction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing devices fail to accurately distinguish between running and walking, and cannot effectively estimate distance covered when a person alternates between these activities, assuming a constant stride length regardless of pace.
Innovation Solution
A method using an accelerometer to monitor acceleration profiles, identifying walking and running steps by detecting fluctuating accelerations, measuring peak values, and distinguishing between the two based on specific thresholds and free-fall moments, allowing for separate evaluation and accurate distance estimation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a pedometer uses an accelerometer to count steps and estimate distance, then step counting function is provided, but it cannot distinguish between walking and running steps leading to inaccurate distance estimation
Solution Approach 1:
The patent applies parameter changes by analyzing different characteristics of acceleration profiles to distinguish between walking and running. Specifically, it examines the frequency, amplitude, and pattern of acceleration fluctuations to identify whether steps are from walking or running, thereby enabling accurate distance estimation for each exercise mode separately.
Solution Approach 2:
The system dynamically adjusts its analysis based on the detected exercise mode. By continuously monitoring acceleration patterns and identifying transitions between walking and running, the system adapts its measurement approach to match the current activity, improving overall measurement precision across varying exercise intensities.
2Ease of operation
If the device assumes constant stride length for all exercises, then calculation is simplified, but distance estimation becomes inaccurate when user alternates between running and walking
Solution Approach 1:
The patent segments the exercise routine into distinct walking and running portions based on acceleration profile analysis. By dividing the continuous movement data into separate exercise mode segments, the system can apply different stride length parameters for each segment, maintaining calculation feasibility while improving accuracy.
Solution Approach 2:
The system dynamically determines stride length based on the detected exercise mode rather than assuming a constant value. When walking is detected, one stride length parameter is applied; when running is detected, a different parameter is used, allowing the calculation to adapt to changing exercise conditions.
3Ease of operation
If the device uses only acceleration magnitude for step detection, then detection is straightforward, but it cannot identify free-fall moments characteristic of running
Solution Approach 1:
The patent extends the analysis from single-dimensional acceleration magnitude to multi-dimensional acceleration profile analysis. By examining the temporal pattern, frequency content, and directional components of acceleration, the system can identify characteristic features of running such as free-fall moments, while maintaining a systematic detection approach.
Solution Approach 2:
The system changes the parameters being analyzed from simple acceleration magnitude to multiple acceleration characteristics including frequency, amplitude modulation, and temporal patterns. This enables identification of running-specific features like the brief free-fall period when both feet are off the ground during running gait.
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
Enables accurate identification and evaluation of walking and running contributions to an exercise routine, reducing error in distance estimation and allowing for more precise tracking of alternating activities.
Implementation Method 1
detecting fluctuating accelerations experienced by a person moving on foot
Implementation Method 2
providing an accelerometer to monitoring the acceleration profile of a person moving on foot
Data Source
AI summary
A person's exercise monitored using an accelerometer to detect and count the steps of the person. Furthermore, acceleration profiling distinguishes walking from running and total distance covered in the exercise is estimated using different stride lengths for walking and running. The person's pace may be maintained using a music playback unit with an adjustable rate of playback to influence the pace of the person with the music tempo.