An Intelligent IoT Sensing System for Rail Vehicle Running States Based on TinyML

Real-time identification of the running state is one of the key technologies for a smart rail vehicle. However, it is a challenge to accurately real-time sense the complex running states of the rail vehicle on an Internet-of-Things (IoT) edge device. Traditional systems usually upload a large amount of real-time data from the vehicle to the cloud for identification, which is laborious and inefficient. In this paper, an intelligent identification method for rail vehicle running state is proposed based on Tiny Machine Learning (TinyML) technology, and an IoT system is developed with small size and low energy consumption. The system uses a Micro-Electro-Mechanical System (MEMS) sensor to collect acceleration data for machine learning training. A neural network model for recognizing the running state of rail vehicles is built and trained by defining a machine learning running state classification model. The trained recognition model is deployed to the IoT edge device at the vehicle side, and an offset time window method is utilized for real-time state sensing. In addition, the sensing results are uploaded to the IoT server for visualization. The experiments on the subway vehicle showed that the system could identify six complex running states in real-time with over 99% accuracy using only one IoT microcontroller. The model with three axes converges faster than the model with one. The model recognition accuracy remained above 98% and 95%, under different installation positions on the rail vehicle and the zero-drift phenomenon of the MEMS acceleration sensor, respectively. The presented method and system can also be extended to edge-aware applications of equipment such as automobiles and ships.

View this article on IEEE Xplore


Video Based Mobility Monitoring of Elderly People Using Deep Learning Models

In recent years, the number of older people living alone has increased rapidly. Innovative vision systems to remotely assess people’s mobility can help healthy, active, and happy aging. In the related literature, the mobility assessment of older people is not yet widespread in clinical practice. In addition, the poor availability of data typically forces the analyses to binary classification, e.g. normal/anomalous behavior, instead of processing exhaustive medical protocols. In this paper, real videos of elderly people performing three mobility tests of a clinical protocol are automatically categorized, emulating the complex evaluation process of expert physiotherapists. Videos acquired using low-cost cameras are initially processed to obtain skeletal information. A proper data augmentation technique is then used to enlarge the dataset variability. Thus, significant features are extracted to generate a set of inputs in the form of time series. Four deep neural network architectures with feedback connections, even aided by a preliminary convolutional layer, are proposed to label the input features in discrete classes or to estimate a continuous mobility score as the result of a regression task. The best results are achieved by the proposed Conv-BiLSTM classifier, which achieves the best accuracy, ranging between 88.12% and 90%. Further comparisons with shallow learning classifiers still prove the superiority of the deep Conv-BiLSTM classifier in assessing people’s mobility, since deep networks can evaluate the quality of test executions.

View this article on IEEE Xplore


Downhole Microseismic Monitoring Using Time-Division Multiplexed Fiber-Optic Accelerometer Array

Microseismic monitoring is of importance for several geoscience research aspects and for applications in oil and gas industry. For signals generated by the ultra-weak microseismic events, conventional moving-coil geophone systems have reached their limit in detection sensitivity especially at high frequency range. Here we for the first time present a specially tailored fiber-optic sensing system targeting at downhole microseismic monitoring. The system contains 30 individual interferometric accelerometers and 2 reference sensors, which are time-division multiplexed into a 12-level vector seismic sensor array. The multiplexed accelerometers can achieve ~50 ng/√Hz noise equivalent acceleration, which is superior to the commercial available moving-coil geophone systems at frequencies above 200 Hz. The measured sensitivity of the accelerometers can reach ~200 rad/g from 10 Hz to 1 kHz. The dynamic range is above 134 dB over the same frequency range and is higher than its electronic counterpart in the low frequency band. Moreover, the sensors can function properly under the harsh condition of 120 °C temperature and 40 MPa pressure over the 4-hour test duration. The sensor array along with the interrogator has been running uninterruptedly over 3 weeks in a multi-stage hydraulic fracturing stimulation field test. On-site results show that our system can clearly resolve the vector nature of both compressional and shear waves generated by the microseismic events.

Published in the IEEE Photonics Society Section within IEEE Access.

View this article on IEEE Xplore

Sensing Methodologies in Agriculture for Soil Moisture and Nutrient Monitoring

Development and deployment of sensing technologies is one of the main steps in achieving sustainability in crop production through precision agriculture. Key sensing methodologies developed for monitoring soil moisture and nutrients with recent advances in the sensing devices reported in literature using those techniques are overviewed in this article. The soil moisture determination has been divided into four main sections describing soil moisture measurement metrics and laboratory-based testing, followed by in-situ, remote and proximal sensing techniques. The application, advantages and limitations for each of the mentioned technologies are discussed. The nutrient monitoring methods are reviewed beginning with laboratory-based methods, ion-selective membrane based sensors, bio-sensors, spectroscopy-based methods, and capillary electrophoresis-based systems for inorganic ion detection. Attention has been given to the core principle of detection while reporting recent sensors developed using the mentioned concepts. The latest works reported on the different sensing methodologies point towards the trend of developing low-cost, easy to use, field-deployable or portable sensing systems aimed towards improving technology adoption in crop production leading to efficient site-specific soil and crop management which in turn will bring us closer to reaching sustainability in the practice of agriculture.

View this article on IEEE Xplore

2018 IEEE Access Best Multimedia Award Part 1 Winners

IEEE Access would like to congratulate the winners of the 2018 IEEE Access Best Multimedia Award Part 1 and recipients of a $500 USD Amazon gift card for their fine contributions to IEEE Access. The full article entitled, “Design and Optimization of a Polar Satellite Mission to Complement the Copernicus Systems” can be found by clicking here.


Most Popular Article of 2017: Internet of Things for Smart Healthcare: Technologies, Challenges, and Opportunities

Internet of Things (IoT) technology has attracted much attention in recent years for its potential to alleviate the strain on healthcare systems caused by an aging population and a rise in chronic illness. Standardization is a key issue limiting progress in this area, and thus this paper proposes a standard model for application in future IoT healthcare systems. This survey paper then presents the state-of-the-art research relating to each area of the model, evaluating their strengths, weaknesses, and overall suitability for a wearable IoT healthcare system. Challenges that healthcare IoT faces including security, privacy, wearability, and low-power operation are presented, and recommendations are made for future research directions.

View this article on IEEE Xplore

SmartCityWare: A Service-Oriented Middleware for Cloud and Fog Enabled Smart City Services

Smart cities are becoming a reality. Various aspects of modern cities are being automated and integrated with information and communication technologies to achieve higher functionality, optimized resources utilization, and management, and improved quality of life for the residents. Smart cities rely heavily on utilizing various software, hardware, and communication technologies to improve the operations in areas, such as healthcare, transportation, energy, education, logistics, and many others, while reducing costs and resources consumption. One of the promising technologies to support such efforts is the Cloud of Things (CoT). CoT provides a platform for linking the cyber parts of a smart city that are executed on the cloud with the physical parts of the smart city, including residents, vehicles, power grids, buildings, water networks, hospitals, and other resources. Another useful technology is Fog Computing, which extends the traditional Cloud Computing paradigm to the edge of the network to enable localized and real-time support for operating-enhanced smart city services. However, proper integration and efficient utilization of CoT and Fog Computing is not an easy task. This paper discusses how the service-oriented middleware (SOM) approach can help resolve some of the challenges of developing and operating smart city services using CoT and Fog Computing. We propose an SOM called SmartCityWare for effective integration and utilization of CoT and Fog Computing. SmartCityWare abstracts services and components involved in smart city applications as services accessible through the service-oriented model. This enhances integration and allows for flexible inclusion and utilization of the various services needed in a smart city application. In addition, we discuss the implementation and experimental issues of SmartCityWare and demonstrate its use through examples of smart city applications.

View this article on IEEE Xplore