Intelligent Posture Monitoring System

With the shift to at-home work due to the Covid-19 pandemic, longer hours are spent sitting in front of a computer without proper ergonomic seating available in most home-office settings. Most home office arrangements often lack the necessary back support needed for prolonged periods of sedentary work. The goal of the proposed system is to automatically track a user's postural positions throughout the day through the use of a non-invasive, wearable system and automatically provide feedback from an algorithm to warn the user to correct or change their poor posture. This is done by placing magnets in the form of a rectangular grid on a shirt as well as an MMR sensor on the chest of the body. The onboard magnetic sensor records the data values from the grid of magnetics, which is then, along with data recorded from the onboard accelerometer, analyzed to determine the position of the user. A trained algorithm recognizes and automatically detects the spinal position of the user from the recorded data points and provides direction to alter their posture. These recommendations act as a warning system and allow the user to self-monitor and correct their own behavior to prevent back and neck pain and reduce the chance of long-lasting damage that can result from poor posture. © 2022 IEEE.

Re-configurable, expandable, and cost-effective heterogeneous FPGA cluster approach for resource-constrained data analysis

Field programmable gate arrays (FPGAs) have become widely prevalent in recent years as a great alternative to application-specific integrated circuits (ASIC) and as a potentially cheap alternative to expensive graphics processing units (GPUs). Introduced as a prototyping solution for ASIC, FPGAs are now widely popular in applications such as artificial intelligence (AI) and machine learning (ML) models that require processing data rapidly. As a relatively low-cost option to GPUs, FPGAs have the advantage of being reprogrammed to be used in almost any data-driven application. In this work, we propose an easily scalable and cost-effective cluster-based co-processing system using FPGAs for ML and AI applications that is easily reconfigured to the requirements of each user application. The aim is to introduce a clustering system of FPGA boards to improve the efficiency of the training component of machine learning algorithms. Our proposed configuration provides an opportunity to utilise relatively inexpensive FPGA development boards to produce a cluster without expert knowledge in VHDL, Verilog, or the system designs related to FPGA development. Consisting of two parts–a computer-based host application to control the cluster and an FPGA cluster connected through a high-speed Ethernet switch, allows the users to customise and adapt the system without much effort. The methods proposed in this paper provide the ability to utilise any FPGA board with an Ethernet port to be used as a part of the cluster and unboundedly scaled. To demonstrate the effectiveness of the proposed work, a two-part experiment to demonstrate the flexibility and portability of the proposed work–a homogeneous and heterogeneous cluster, was conducted with results compared against a desktop computer and combinations of FPGAs in two clusters. Data sets ranging from 60,000 to 14 million, including stroke prediction and covid-19, were used in conducting the experiments. Results suggest that the proposed system in this work performs close to 70% faster than a traditional computer with similar accuracy rates. © 2022 Informa UK Limited, trading as Taylor & Francis Group.

Detection and Monitoring of High Germ Spreading Activities Using Wearable Sensors

This paper discusses our preliminary study and results on data collection, data processing, feature extraction and classification in developing a machine learning model for detecting daily activities of living, especially the germ spreading activities during the COVID-19 pandemic. In this research, a Mbient Lab MetaWear wearable sensor system is used to collect arm and hand motion data from subjects performing various activities. After data was collected from these different activities, the data was processed. Important statistical time-domain features and frequency domain features, such as the total energy in different frequency bands, were extracted with respect to these different activities to differentiate between them. Various features were collected to create a feature matrix and were used to train different Machine Learning algorithms to determine the germ spreading activity classification accuracy. Using the ensemble bagged tree model, a classification accuracy of 97.0% was obtained. © 2021 IEEE.

Magnet Integrated Shirt for Upper Body Posture Detection Using Wearable Magnetic Sensors

COVID-19 is a global pandemic that has caused an increase in remote work. Sitting in various positions at home without the proper back support is undesirable and can cause chronic back pain and other undesired side effects. Therefore, a new non-intrusive method to continuously monitor back postures in homes is proposed. A shirt is designed with integrated magnets. A magnetic sensor would be placed above the body's sternum, and magnets will be implemented on a shirt. The sensor will help ascertain the back posture position (straight or curved) and help provide feedback to mend the posture if deformed. In this paper, the initial results using the proposed system are presented using a wearable sensor system with a magnet integrated garment that can continuously monitor the varying sitting postures throughout daily lives. In addition, we discuss how the lower body posture affects the magnetic recording otherwise not detectable using the accelerometer-based systems currently available on the market.

Detection of High Germ Spreading Activities

This paper discusses the preliminary process of data collection, data processing, and feature extraction and selection in applications of developing a machine learning model for activity detection, especially the germ spreading activities during the COVID-19 pandemic. In this research, a MetaWear wearable device is used to collect arm and hand motion data from a subject performing various activities. After data was collected from these different activities, the data was processed, and important time-domain features as well as frequency domain features, such as the total energy contained in different frequency bands, were extracted in respect to these different activities with the objective of differentiating between these various activities. Various features were collected to create a feature matrix and input to different Machine Learning algorithms to determine the classification accuracy of the germ spreading activities. Using the ensemble bagged tree model, a classification accuracy of 99.4% was obtained.