Accelerometer Speed Estimation via Frequency Spectrum Analysis

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current solutions for determining the speed of movement and pose of communication devices assume fixed sensor orientations and placements, which do not align with typical user behavior, leading to inefficiencies in context-aware behavior services.

Innovation Solution

The method employs accelerometer-based mechanisms using regularized kernel techniques, such as regularized least squares and support vector machines, to estimate speed and classify device poses without assuming fixed sensor positions, transforming acceleration data into frequency components and comparing them to training spectrums for accurate speed and pose determination.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If fixed sensor orientations and placements are assumed, then device complexity is reduced, but adaptability to various user movements and poses deteriorates

Engineering Contradiction:

Improvesensor placement complexityVSAvoidadaptability to user movements

Core Design Contradiction:

Device complexityVSAdaptability or versatility

Solution Approach 1:

The system dynamically adapts to different device poses and user movements by continuously analyzing accelerometer data patterns rather than assuming fixed sensor orientations. The method classifies device poses in real-time and adjusts speed estimation algorithms accordingly, making the system flexible and adaptive to various usage scenarios without requiring fixed sensor placement

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system changes parameters of the acceleration data representation by transforming time-domain accelerometer readings into frequency-domain characteristics through spectral analysis. By extracting frequency components and comparing them to training spectrums, the system can accurately determine speed and pose regardless of fixed sensor orientation assumptions

Inventive Principle:

Principle #35Parameter changes

2Adaptability or versatility

If accelerometer-based mechanisms without fixed sensor assumptions are used, then adaptability to various user movements improves, but device complexity increases

Engineering Contradiction:

Improveadaptability to user movementsVSAvoidprocessing complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The system performs preliminary action by collecting and storing training data with labeled device poses and speed information before actual operation. This pre-processing creates reference spectrums that simplify real-time analysis, allowing the system to quickly compare current accelerometer patterns against known patterns without complex real-time calculations

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system replaces complex mechanical sensor placement requirements with signal processing techniques. Instead of relying on precise physical sensor positioning, the method uses mathematical transformations of accelerometer data and pattern recognition algorithms to achieve pose and speed determination, substituting mechanical precision with computational analysis

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Measurement precision

If fixed sensor positions are assumed, then measurement precision may be maintained under controlled conditions, but reliability across diverse usage scenarios deteriorates

Engineering Contradiction:

Improvespeed estimation precisionVSAvoidreliability across usage scenarios

Core Design Contradiction:

Measurement precisionVSReliability

Solution Approach 1:

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system uses feedback from classified device poses to improve speed estimation accuracy. By continuously monitoring accelerometer patterns, classifying the current pose, and adjusting processing parameters based on the classified pose type, the system creates a closed-loop system that maintains precision across diverse usage conditions

Inventive Principle:

Principle #23Feedback

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

This approach achieves high predictive performance with speed estimation errors less than 8% and device pose accuracy of about 95%, effectively addressing the limitations of existing solutions by adapting to various user movements and poses.

Implementation Method 1

utilizing time-series acceleration data received from a sensor such as, for example, an accelerometer

Methodology Applied

Scientific EffectAccelerometer: Accelerometer

Data Source

PatentUS9069003B2Methods, apparatuses and computer program products for determining speed of movement of a device and device pose classification

Publication Date: 2015.06.30 NOKIA CORP

AI summary

An apparatus for determining a speed of cyclic motion of a device or user and one or more poses of a device may include a processor and memory storing executable computer code causing the apparatus to at least perform operations including receiving one or more determined acceleration values during one or more time periods in response to detected cyclic motion of a user moving with an apparatus. The computer program code may further cause the apparatus to transform the acceleration values to one or more corresponding frequency components associated with the acceleration values. The computer program code may further cause the apparatus to determine a speed of the cyclic motion of the user based in part on comparing a spectrum of the frequency components to one or more spectrums of distribution associated with respective one or more speeds of training data. Corresponding methods and computer program products are also provided.