Wearable Sensor Classification With Edge Offloading for Low Power
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Computing devices, particularly wearable devices, face challenges with high energy consumption, heat generation, and short battery life due to energy-intensive operations like image and audio processing, which are uncomfortable to wear for extended periods.
Innovation Solution
A distributed sensor data processing system where low-power wearable devices offload energy-intensive operations to connected computing devices or server computers via wireless connections, using smaller machine learning models for initial detection and transmitting data for more complex processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If energy-intensive operations (image and audio processing) are performed on wearable devices, then computing performance is improved, but power consumption increases and battery life decreases
Solution Approach 1:
The processing system is segmented into two parts: a lightweight wearable device for basic operations and a remote computing device for intensive operations. The wearable device performs simple classification tasks while offloading complex processing to the remote device, thereby reducing local power consumption while maintaining overall system performance.
Solution Approach 2:
A wireless communication interface acts as an intermediary between the wearable device and the remote computing device. This mediator enables the transfer of data and computational tasks, allowing the wearable device to leverage remote computing power without requiring direct integration of heavy processing components.
2Productivity
If energy-intensive operations are performed on wearable devices, then computing capability is improved, but heat generation increases making the device uncomfortable
Solution Approach 1:
The heat-generating intensive processing operations are extracted from the wearable device and relocated to a remote computing device. By removing the heavy computational workload from the wearable device, the temperature and discomfort issues are eliminated while the wearable device maintains its computing capabilities through remote processing.
3Productivity
If more circuit components are added to wearable devices for processing capabilities, then computing performance is improved, but device weight increases
Solution Approach 1:
The heavy circuit components required for intensive image and audio processing are extracted from the wearable device. By removing these components and performing processing remotely, the wearable device achieves reduced weight while maintaining computing performance through the distributed system architecture.
4Duration of action of moving object
If battery capacity is increased to extend usage time, then device autonomy is improved, but device weight and size increase
Solution Approach 1:
The battery capacity requirement is extracted and reduced by offloading intensive processing tasks. Since the wearable device performs only lightweight operations locally, the required battery capacity is significantly reduced, thereby extending usage time without increasing weight.
Data Source
AI summary
According to an aspect, a method for distributed sound/image recognition using a wearable device includes receiving, via at least one sensor device, sensor data, and detecting, by a classifier of the wearable device, whether or not the sensor data includes an object of interest. The classifier configured to execute a first machine learning (ML) model. The method includes transmitting, via a wireless connection, the sensor data to a computing device in response to the object of interest being detected within the sensor data, where the sensor data is configured to be used by a second ML model on the computing device or a server computer for further sound/image classification.