skip to main content
US FlagAn official website of the United States government
dot gov icon
Official websites use .gov
A .gov website belongs to an official government organization in the United States.
https lock icon
Secure .gov websites use HTTPS
A lock ( lock ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

Explore Research Products in the PAR
It may take a few hours for recently added research products to appear in PAR search results.


Title: A Review of Embedded Machine Learning Based on Hardware, Application, and Sensing Scheme
Machine learning is an expanding field with an ever-increasing role in everyday life, with its utility in the industrial, agricultural, and medical sectors being undeniable. Recently, this utility has come in the form of machine learning implementation on embedded system devices. While there have been steady advances in the performance, memory, and power consumption of embedded devices, most machine learning algorithms still have a very high power consumption and computational demand, making the implementation of embedded machine learning somewhat difficult. However, different devices can be implemented for different applications based on their overall processing power and performance. This paper presents an overview of several different implementations of machine learning on embedded systems divided by their specific device, application, specific machine learning algorithm, and sensors. We will mainly focus on NVIDIA Jetson and Raspberry Pi devices with a few different less utilized embedded computers, as well as which of these devices were more commonly used for specific applications in different fields. We will also briefly analyze the specific ML models most commonly implemented on the devices and the specific sensors that were used to gather input from the field. All of the papers included in this review were selected using Google Scholar and published papers in the IEEExplore database. The selection criterion for these papers was the usage of embedded computing systems in either a theoretical study or practical implementation of machine learning models. The papers needed to have provided either one or, preferably, all of the following results in their studies—the overall accuracy of the models on the system, the overall power consumption of the embedded machine learning system, and the inference time of their models on the embedded system. Embedded machine learning is experiencing an explosion in both scale and scope, both due to advances in system performance and machine learning models, as well as greater affordability and accessibility of both. Improvements are noted in quality, power usage, and effectiveness.  more » « less
Award ID(s):
1652944 2015573
PAR ID:
10417386
Author(s) / Creator(s):
;
Date Published:
Journal Name:
Sensors
Volume:
23
Issue:
4
ISSN:
1424-8220
Page Range / eLocation ID:
2131
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. ; (, Proceedings of the Twenty-Ninth International Joint Conference on Artificial Intelligence)
    null (Ed.)
    With the recent advances in both machine learning and embedded systems research, the demand to deploy computational models for real-time execution on edge devices has increased substantially. Without deploying computational models on edge devices, the frequent transmission of sensor data to the cloud results in rapid battery draining due to the energy consumption of wireless data transmission. This rapid power dissipation leads to a considerable reduction in the battery lifetime of the system, therefore jeopardizing the real-world utility of smart devices. It is well-established that for difficult machine learning tasks, models with higher performance often require more computation power and thus are not power-efficient choices for deployment on edge devices. However, the trade-offs between performance and power consumption are not well studied. While numerous methods (e.g., model compression) have been developed to obtain an optimal model, these methods focus on improving the efficiency of a single model. In an entirely new direction, we introduce an effective method to find a combination of multiple models that are optimal in terms of power-efficiency and performance by solving an optimization problem in which both performance and power consumption are taken into account. Experimental results demonstrate that on the ImageNet dataset, we can achieve a 20% energy reduction with only 0.3% accuracy drop compared to Squeeze-and-Excitation Networks. Compared to a pruned convolutional neural network for human activity recognition, while consuming 1.7% less energy, our proposed policy achieves 1.3% higher accuracy. 
    more » « less
  2. ; ; ; ; (, IEEE 2022 International Symposium on Medical Robotics (ISMR))
    Detection of the user’s walking is a critical part of exoskeleton technology for the full automation of smooth and seamless assistance during movement transitions. Researchers have taken several approaches in developing a walk detection system by using different kinds of sensors; however, only a few solutions currently exist which can detect these transitions using only the sensors embedded on a robotic hip exoskeleton (i.e., hip encoders and a trunk IMU), which is a critical consideration for implementing these systems in-the-loop of a hip exoskeleton controller. As a solution, we explored and developed two walk detection models that implemented a finite state machine as the models switched between walking and standing states using two transition conditions: stand-to-walk and walk-to-stand. One of our models dynamically detected the user’s gait cycle using two hip encoders and an IMU; the other model only used the two hip encoders. Our models were developed using a publicly available dataset and were validated online using a wearable sensor suite that contains sensors commonly embedded on robotic hip exoskeletons. The two models were then compared with a foot contact estimation method, which served as a baseline for evaluating our models. The results of our online experiments validated the performance of our models, resulting in 274 ms and 507 ms delay time when using the HIP+IMU and HIP ONLY model, respectively. Therefore, the walk detection models established in our study achieve reliable performance under multiple locomotive contexts without the need for manual tuning or sensors additional to those commonly implemented on robotic hip exoskeletons. 
    more » « less
  3. ; ; ; ; (, Sensors)
    Rapid growth of sensors and the Internet of Things is transforming society, the economy and the quality of life. Many devices at the extreme edge collect and transmit sensitive information wirelessly for remote computing. The device behavior can be monitored through side-channel emissions, including power consumption and electromagnetic (EM) emissions. This study presents a holistic self-testing approach incorporating nanoscale EM sensing devices and an energy-efficient learning module to detect security threats and malicious attacks directly at the front-end sensors. The built-in threat detection approach using the intelligent EM sensors distributed on the power lines is developed to detect abnormal data activities without degrading the performance while achieving good energy efficiency. The minimal usage of energy and space can allow the energy-constrained wireless devices to have an on-chip detection system to predict malicious attacks rapidly in the front line. 
    more » « less
  4. ; (, IEEE)
    While machine learning models perform well on offline data, assessing their performance in real-world, resource-constrained environments-considering accuracy, prediction time, power consumption, and memory usage-is crucial for practical applications. This research implements a mobile-based Human Activity Recognition solution to classify three postures-sitting, standing, and walking-using smartphone sensors, specifically accelerometer, gyroscope, and magnetometer. Time-domain features extracted from these sensors were used, with Random Forest employed for feature selection. One traditional machine learning model, Logistic Regression, and one deep learning model, Convolutional Neural Network, were trained and deployed via an Android application for real-time evaluation. While the Convolutional Neural Network achieved higher accuracy and better memory efficiency, Logistic Regression demonstrated faster prediction times during real-time use. Both models showed reduced accuracy for standing and walking postures in real-world conditions, emphasizing the challenges of deploying machine learning models in dynamic environments. This study highlights the importance of evaluating machine learning models in real-world settings to ensure reliability and efficiency, particularly in resource-constrained environments. 
    more » « less
  5. ; ; (, IEEE sensors journal)
    Sitting is the most common status of modern human beings. Some sitting postures may bring health issues. To prevent the harm from bad sitting postures, a local sitting posture recognition system is desired with low power consumption and low computing overhead. The system should also provide good user experience with accuracy and privacy. This paper reports a novel posture recognition system on an office chair that can categorize seven different health-related sitting postures. The system uses six flex sensors, an Analog to Digital Converter (ADC) board and a Machine Learning algorithm of a two-layer Artificial Neural Network (ANN) implemented on a Spartan-6 Field Programmable Gate Array (FPGA). The system achieves 97.78% accuracy with a floating-point evaluation and 97.43% accuracy with the 9-bit fixed-point implementation. The ADC control logic and the ANN are constructed with a maximum propagation delay of 8.714 ns. The dynamic power consumption is 7.35 mW when the sampling rate is 5 Sample/second with the clock frequency of 5 MHz. 
    more » « less
  • Citation Formats
  • MLA
  • APA
  • Chicago
  • BibTeX
  • Save / Share this Record
  • Send to Email