Activity Tracking System Using Segmented Parallel Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing activity tracking devices struggle to accurately sense, record, and analyze all types of user activities, such as walking, running, cycling, and others, due to variations in sensor data and how the device is carried, leading to classification challenges.
Innovation Solution
A system that uses a mobile communication device with movement sensors and a wireless interface, communicating with computing hardware to analyze sensor signals, classify them into temporal zones, and determine likely activity types, with user confirmation for improving algorithm accuracy, employing supervised or semisupervised classification algorithms and parallel processing for enhanced accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing tracking devices use basic sensor data collection, then device simplicity is maintained, but activity classification accuracy deteriorates due to inability to distinguish between similar activities like walking vs running
Solution Approach 1:
The patent segments the activity classification process into multiple independent processor instances, where each processor is dedicated to identifying characteristics of sensor data corresponding to a specific activity type. This segmentation allows the system to achieve high classification accuracy by specializing each processor for a particular activity while maintaining overall system manageability.
Solution Approach 2:
The patent employs a universal sensor system (accelerometer, GPS receiver, proximity sensor) that serves multiple functions for detecting various types of user activities. These same sensors are used across all activity detection processors, allowing the system to maintain simplicity in hardware while achieving versatility and accuracy in activity classification through software-based multi-functionality.
2Measurement precision
If the system processes sensor data in real-time with high accuracy, then activity identification precision is improved, but processing time and computational resources increase
Solution Approach 1:
The patent divides the computational workload into multiple parallel processor instances, each handling a specific activity type independently. This segmentation enables simultaneous processing of different activity detections without sequential delays, reducing overall processing time while maintaining high precision for each activity type through specialized processing.
Solution Approach 2:
The system performs preliminary classification of sensor data into temporal zones before detailed activity analysis. By pre-organizing data into time-based segments and identifying characteristic patterns in advance, the system reduces the computational complexity of real-time analysis, enabling faster processing without sacrificing detection precision.
3Productivity
If the system uses multiple processors for parallel processing, then productivity of activity analysis is improved, but device complexity increases
Solution Approach 1:
The patent implements multiple processor instances that are segmented by function, with each processor dedicated to identifying characteristics corresponding to a specific activity type. This functional segmentation increases analysis throughput by enabling parallel processing while keeping each individual processor simple and manageable, avoiding the complexity of a single monolithic processing system.
Solution Approach 2:
The system uses identical copies of the processor architecture for each activity type, where each copy follows the same structural template but is specialized for a particular activity. This copying approach standardizes the processing system, making it easier to manage and maintain while achieving high productivity through parallel operation of multiple identical units.
4Measurement precision
If the system collects comprehensive sensor data from multiple sources, then measurement completeness is improved, but data processing complexity and energy consumption increase
Solution Approach 1:
The patent extracts and processes only the specific sensor data characteristics that are relevant to each activity type, rather than processing all sensor data comprehensively. Each processor instance extracts only the characteristic features needed for its specific activity detection, reducing overall processing complexity and energy consumption while maintaining complete activity sensing through the collective effort of multiple specialized processors.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present disclosure provides a system for tracking and recording movements of a mobile communication device that includes one or more movement sensors. The mobile communication device communicates sensor signals to the system, wherein the sensor signals are indicative of motion associated with activities to which the mobile communication device is exposed by its user. The sensor signals are pre-classified to generate the intermediate data, where the intermediate data is processed in one or more processors to generate indications of likely activities associated with the sensor signals, and an aggregate of the indications is computed to provide an analysis of activities associated with the sensor signals. The system analyses sensor signals to classify them into one or more temporal zones and determines a most likely activity type associated with a given temporal zone. The information indicating most likely activity types is sent to the mobile communication device.