Sensor-Based Indoor Fire Forecasting Using Transformer Encoder

Indoor fires may cause casualties and property damage, so it is important to develop a system that predicts fires in advance. There have been studies to predict potential fires using sensor values, and they mostly exploited machine learning models or recurrent neural networks. In this paper, we propose a stack of Transformer encoders for fire prediction using multiple sensors. Our model takes the time-series values collected from the sensors as input, and predicts the potential fire based on the sequential patterns underlying the time-series data. We compared our model with traditional machine learning models and recurrent neural networks on two datasets. For a simple dataset, we found that the machine learning models are better than ours, whereas our model gave better performance for a complex dataset. This implies that our model has a greater potential for real-world applications that probably have complex patterns and scenarios.

A Hybrid Convolutional and Recurrent Neural Network for Multi-Sensor Pile Damage Detection with Time Series

Machine learning (ML) algorithms are increasingly applied to structure health monitoring (SHM) problems. However, their application to pile damage detection (PDD) is hindered by the complexity of the problem. A novel multi-sensor pile damage detection (MSPDD) method is proposed in this paper to extend the application of ML algorithms in the automatic identification of PDD. The time-series signals collected by multiple sensors during the pile integrity test are first processed by the traveling wave decomposition (TWD) theory and are then input into a hybrid one-dimensional (1D) convolutional and recurrent neural network. The hybrid neural network can achieve the automatic multi-task identification of pile damage detection based on the time series of MSPDD results. Finally, the analytical solution-based sample set is utilized to evaluate the performance of the proposed hybrid model. The outputs of the multi-task learning framework can provide a detailed description of the actual pile quality and provide strong support for the classification of pile quality as well.

STASISM: A Versatile Serious Gaming Multi-Sensor Platform for Personalized Telerehabilitation and Telemonitoring

Telemonitoring and telerehabilitation have shown promise in delivering individualized healthcare remotely. We introduce STASISM, a sensor-based telerehabilitation and telemonitoring system, in this work. This platform has been created to facilitate individualized telerehabilitation and telemonitoring for those who need rehabilitation or ongoing monitoring. To gather and analyze pertinent and validated physiological, kinematic, and environmental data, the system combines a variety of sensors and data analytic methodologies. The platform facilitates customized rehabilitation activities based on individual needs, allows for the remote monitoring of a patient's progress, and offers real-time feedback. To protect the security of patient data and to safeguard patient privacy, STASISM also provides secure data transmission and storage. The platform has the potential to significantly improve the accessibility and efficacy of telerehabilitation and telemonitoring programs, enhancing patients' quality of life and allowing healthcare professionals to provide individualized care outside of traditional clinical settings.

An Improved ResNet-1d with Channel Attention for Tool Wear Monitor in Smart Manufacturing

Tool wear is a key factor in the machining process, which affects the tool life and quality of the machined work piece. Therefore, it is crucial to monitor and diagnose the tool condition. An improved CaAt-ResNet-1d model for multi-sensor tool wear diagnosis was proposed. The ResNet18 structure based on a one-dimensional convolutional neural network is adopted to make the basic model architecture. The one-dimensional convolutional neural network is more suitable for feature extraction of time series data. Add the channel attention mechanism of CaAt1 to the residual network block and the channel attention mechanism of CaAt5 automatically learns the features of different channels. The proposed method is validated on the PHM2010 dataset. Validation results show that CaAt-ResNet-1d can reach 89.27% accuracy, improving by about 7% compared to Gated-Transformer and 3% compared to Resnet18. The experimental results demonstrate the capacity and effectiveness of the proposed method for tool wear monitor.

A Multi-Sensor Environmental Perception System for an Automatic Electric Shovel Platform

Electric shovels have been widely used in heavy industrial applications, such as mineral extraction. However, the performance of the electric shovel is often affected by the complicated working environment and the proficiency of the operator, which will affect safety and efficiency. To improve the extraction performance, it is particularly important to study an intelligent electric shovel with autonomous operation technology. An electric shovel experimental platform for intelligent technology research and testing is proposed in this paper. The core of the designed platform is an intelligent environmental sensing/perception system, in which multiple sensors, such as RTK (real-time kinematic), IMU (inertial measurement unit) and LiDAR (light detection and ranging), have been employed. By appreciating the multi-directional loading characteristics of electric shovels, two 2D-LiDARs have been used and their data are synchronized and fused to construct a 3D point cloud. The synchronization is achieved with the assistance of RTK and IMU, which provide pose information of the shovel. In addition, in order to down-sample the LiDAR point clouds to facilitate more efficient data analysis, a new point cloud data processing algorithm including a bilateral-filtering based noise filter and a grid-based data compression method is proposed. The designed platform, together with its sensing system, was tested in different outdoor environment conditions. Compared with the original LiDAR point cloud, the proposed new environment sensing/perception system not only guarantees the characteristic points and effective edges of the measured objects, but also reduces the amount of processing point cloud data and improves system efficiency. By undertaking a large number of experiments, the overall measurement error of the proposed system is within 50 mm, which is well beyond the requirements of electric shovel application. The environment perception system for the automatic electric shovel platform has great research value and engineering significance for the improvement of the service problem of the electric shovel.

CNN-Based Volume Flow Rate Prediction of Oil-Gas-Water Three-Phase Intermittent Flow from Multiple Sensors

In this paper, we propose a deep-learning-based method using a convolutional neural network (CNN) to predict the volume flow rates of individual phases in the oil-gas-water three-phase intermittent flow simultaneously by analyzing the measurement data from multiple sensors, including a temperature sensor, a pressure sensor, a Venturi tube and a microwave sensor. To build datasets, a series of experiments for the oil-gas-water three-phase intermittent flow in a horizontal pipe, in which gas volume fraction and water-in-liquid ratio ranges are 23.77-94.45% and 14.95-86.97%, respectively, and gas flow superficial velocity and liquid flow superficial velocity ranges are 0.66-5.23 and 0.27-2.14 m/s, respectively, have been carried out on a test loop pipeline. The preliminary results indicate that the model can provide relative prediction errors on the testing-1 dataset for the volume flow rates of oil-phase, gas-phase and water-phase within ±10% with 94.49%, 92.56% and 95.71% confidence levels, respectively. Additionally, the prediction results on the testing-2 dataset also demonstrate the generalization ability of the model. The consuming time of a prediction with one sample is 0.43 s on an Intel Xeon CPU E5-2678 v3, and 0.01 s on an NVIDIA GeForce GTX 1080 Ti GPU. Hence, the proposed CNN-based prediction model, which can fulfill the real-time application requirements in the petroleum industry, reveals the potential of using deep learning to obtain accurate results in the multiphase flow measurement field.

Evaluation for Bearing Wear States Based on Online Oil Multi-Parameters Monitoring

As bearings are critical components of a mechanical system, it is important to characterize their wear states and evaluate health conditions. In this paper, a novel approach for analyzing the relationship between online oil multi-parameter monitoring samples and bearing wear states has been proposed based on an improved gray k-means clustering model (G-KCM). First, an online monitoring system with multiple sensors for bearings is established, obtaining oil multi-parameter data and vibration signals for bearings through the whole lifetime. Secondly, a gray correlation degree distance matrix is generated using a gray correlation model (GCM) to express the relationship of oil monitoring samples at different times and then a KCM is applied to cluster the matrix. Analysis and experimental results show that there is an obvious correspondence that state changing coincides basically in time between the lubricants' multi-parameters and the bearings' wear states. It also has shown that online oil samples with multi-parameters have early wear failure prediction ability for bearings superior to vibration signals. It is expected to realize online oil monitoring and evaluation for bearing health condition and to provide a novel approach for early identification of bearing-related failure modes.

Highly Integrated MEMS-ASIC Sensing System for Intracorporeal Physiological Condition Monitoring

In this paper, a highly monolithic-integrated multi-modality sensor is proposed for intracorporeal monitoring. The single-chip sensor consists of a solid-state based temperature sensor, a capacitive based pressure sensor, and an electrochemical oxygen sensor with their respective interface application-specific integrated circuits (ASICs). The solid-state-based temperature sensor and the interface ASICs were first designed and fabricated based on a 0.18-μm 1.8-V CMOS (complementary metal-oxide-semiconductor) process. The oxygen sensor and pressure sensor were fabricated by the standard CMOS process and subsequent CMOS-compatible MEMS (micro-electromechanical systems) post-processing. The multi-sensor single chip was completely sealed by the nafion, parylene, and PDMS (polydimethylsiloxane) layers for biocompatibility study. The size of the compact sensor chip is only 3.65 mm × 1.65 mm × 0.72 mm. The functionality, stability, and sensitivity of the multi-functional sensor was tested ex vivo. Cytotoxicity assessment was performed to verify that the bio-compatibility of the device is conforming to the ISO 10993-5:2009 standards. The measured sensitivities of the sensors for the temperature, pressure, and oxygen concentration are 10.2 mV/°C, 5.58 mV/kPa, and 20 mV·L/mg, respectively. The measurement results show that the proposed multi-sensor single chip is suitable to sense the temperature, pressure, and oxygen concentration of human tissues for intracorporeal physiological condition monitoring.

Combined Dynamic Time Warping with Multiple Sensors for 3D Gesture Recognition

Cyber-physical systems, which closely integrate physical systems and humans, can be applied to a wider range of applications through user movement analysis. In three-dimensional (3D) gesture recognition, multiple sensors are required to recognize various natural gestures. Several studies have been undertaken in the field of gesture recognition; however, gesture recognition was conducted based on data captured from various independent sensors, which rendered the capture and combination of real-time data complicated. In this study, a 3D gesture recognition method using combined information obtained from multiple sensors is proposed. The proposed method can robustly perform gesture recognition regardless of a user’s location and movement directions by providing viewpoint-weighted values and/or motion-weighted values. In the proposed method, the viewpoint-weighted dynamic time warping with multiple sensors has enhanced performance by preventing joint measurement errors and noise due to sensor measurement tolerance, which has resulted in the enhancement of recognition performance by comparing multiple joint sequences effectively.

Noise Correlation Effect on Detection: Signals in Equicorrelated or Autoregressive(1) Gaussian

In this letter, we consider the effect of noise correlation on the error performance of binary hypothesis signal detection, when one of two deterministic signals is received in correlated Gaussian noise. For the likelihood ratio detection scheme, analytical performance results are derived for equicorrelated and autoregressive order one models. Although it is known previously that the best signal lies in the direction of eigenvector corresponding to the minimum eigenvalue of the noise covariance matrix, our investigation of the variation of mean signal-to-noise power ratio as a function of correlation parameter (i) shows how correlation leads to increased probability of error up to a point, beyond which monotonic decrease in error probability with increasing correlation is possible and (ii) provides a max-min signal design solution for the unknown correlation parameter case. Numerical results are also included for some specific signals.

A Portable Farmland Information Collection System with Multiple Sensors

Precision agriculture is the trend of modern agriculture, and it is also one of the important ways to realize the sustainable development of agriculture. In order to meet the production requirements of precision agriculture-efficient use of agricultural resources, and improving the crop yields and quality-some necessary field information in crop growth environment needs to be collected and monitored. In this paper, a farmland information collection system is developed, which includes a portable farmland information collection device based on STM32 (a 32-bit comprehensive range of microcontrollers based on ARM Crotex-M3), a remote server and a mobile phone APP. The device realizes the function of portable and mobile collecting of multiple parameters farmland information, such as chlorophyll content of crop leaves, air temperature, air humidity, and light intensity. UM220-III (Unicore Communication Inc., Beijing, China) is used to realize the positioning based on BDS/GPS (BeiDou Navigation Satellite System, BDS/Global Positioning System, GPS) dual-mode navigation and positioning system, and the CDMA (Code Division Multiple Access, CDMA) wireless communication module is adopted to realize the real-time remote transmission. The portable multi-function farmland information collection system is real-time, accurate, and easy to use to collect farmland information and multiple information parameters of crops.

A method based on multi-sensor data fusion for fault detection of planetary gearboxes

Studies on fault detection and diagnosis of planetary gearboxes are quite limited compared with those of fixed-axis gearboxes. Different from fixed-axis gearboxes, planetary gearboxes exhibit unique behaviors, which invalidate fault diagnosis methods that work well for fixed-axis gearboxes. It is a fact that for systems as complex as planetary gearboxes, multiple sensors mounted on different locations provide complementary information on the health condition of the systems. On this basis, a fault detection method based on multi-sensor data fusion is introduced in this paper. In this method, two features developed for planetary gearboxes are used to characterize the gear health conditions, and an adaptive neuro-fuzzy inference system (ANFIS) is utilized to fuse all features from different sensors. In order to demonstrate the effectiveness of the proposed method, experiments are carried out on a planetary gearbox test rig, on which multiple accelerometers are mounted for data collection. The comparisons between the proposed method and the methods based on individual sensors show that the former achieves much higher accuracies in detecting planetary gearbox faults.