Search

Material classification is pivotal across materials science, engineering, and various industrial sectors. Despite the high accuracy of traditional material classification methods, they often entail large, intricate, and costly setups that demand skilled operators. In this study, we introduce the MCT-array, a newly developed compact RF antenna array system measuring $100\times 100 \times 2$ mm, which functions as a transceiver. This device, equipped with 32 receiving antennas and 2 transmitters, leverages dynamic power adjustments to refine material detection accuracy. The study evaluates three machine learning classifiers, namely Support Vector Machine (SVM), Multi-Layer Perceptron (MLP), and Random Forest (RandF) on twelve different materials. MATLAB simulations are initially conducted to identify optimal transceiver configurations. Following the identification of optimal parameters from these simulations, real-world experiments are conducted with the materials positioned 30 cm away from the antenna. Results demonstrate that RandF achieves a material classification accuracy of 94.84%, followed by SVM at 94.5%, and MLP at 94.1%. Detailed analysis further reveals that RandF is the preferred option for tasks demanding the highest levels of accuracy, SVM strikes an optimal balance between processing speed and accuracy, while MLP stands out for its rapid prediction times, making it especially suitable for real-time applications. Integrating an innovative portable RF transceiver with these machine learning models, achieving an impressive average accuracy of over 94%, represents a scalable and effective solution. This innovation holds significant promise for sectors engaged in material classification, particularly in the realms of robotics and automation.

The real-time and accurate detection of premature ventricular contractions (PVC) in patients is of great significance for preventing the occurrence of high-risk events such as sudden cardiac death and guiding cardiac surgical procedures such as radiofrequency ablation. To improve the diagnostic accuracy and real-time performance, and expand application scenarios, an economical wearable PVC real-time auxiliary diagnosis system based on the multi-parameter squeeze excitation residual network (MP-SE-ResNet) is proposed. We have realized the real-time acquisition, processing, and wireless transmission of dynamic ECGs based on ESP32, furthermore, realized the PVCs recognition based on MP-SE-ResNet. Using the lead-II ECGs in the MIT-BIH arrhythmia databases as training samples, and the network was evaluated using the remainder of this dataset and data recorded by our device, respectively. The accuracy of the MIT-BIH dataset reached 99.34%, and the sensitivity and specificity of PVC recognition reached 98.26% and 99.64%, respectively. Using the ECGs recorded by our system, we achieved the following results: the accuracy was 94.07%, the sensitivity and specificity of PVC were 92.76% and 97.63%, respectively. The experimental results show that the system meets the requirements of remote monitoring and auxiliary diagnosis. Therefore, it provides a new method and design idea for wearable remote arrhythmia monitoring and auxiliary diagnosis.

In the human activity recognition (HAR) application domain, the use of deep learning (DL) algorithms for feature extractions and training purposes delivers significant performance improvements with respect to the use of traditional machine learning (ML) algorithms. However, this comes at the expense of more complex and demanding models, making harder their deployment on constrained devices traditionally involved in the HAR process. The efficiency of DL deployment is thus yet to be explored. We thoroughly investigated the application of TensorFlow Lite simple conversion, dynamic, and full integer quantization compression techniques. We applied those techniques not only to convolutional neural networks (CNNs), but also to long short-term memory (LSTM) networks, and a combined version of CNN and LSTM. We also considered two use case scenarios, namely cascading compression and stand-alone compression mode. This paper reports the feasibility of deploying deep networks onto an ESP32 device, and how TensorFlow compression techniques impact classification accuracy, energy consumption, and inference latency. Results show that in the cascading case, it is not possible to carry out the performance characterization. Whereas in the stand-alone case, dynamic quantization is recommended because yields a negligible loss of accuracy. In terms of power efficiency, both dynamic and full integer quantization provide high energy saving with respect to the uncompressed models: between 31% and 37% for CNN networks, and up to 45% for LSTM networks. In terms of inference latency, dynamic and full integer quantization provide comparable performance.

The idea of the Internet of Things (IoT) is utilized to increase the advantages of internet connectivity. As long as the electrical device is still connected to the internet, IoT can be utilized to operate it. This study applies the concept of IoT as a fan control and monitoring of the humidity and room temperature through a mobile-based application, so that the room can be controlled remotely using a smartphone before use. It is necessary to always keep the room comfortable to use, so that users can increase productivity and always avoid potential disease attacks when the temperature is unstable. In addition, the IoT concept also makes it easier to turn off the fan if at any time you forget to turn it off. This fan is integrated with the ESP32 microcontroller which has been equipped with a Wireless Fidelity (WiFi) module to access data changes in the Firebase Realtime Database. Room temperature and humidity are measured with a DHT22 sensor and processed using Tsukamoto’s Fuzzy Inference System to produce the appropriate fan speed in the Duty Cycle unit. Tools and applications can work as predicted based on the findings of the research conducted. However, the impact of the fan on a closed room cannot cool the room temperature but the room temperature moves up with a change of 0.3 °C to 0.5 °C within 40–75 minutes. Therefore, in the next study it is recommended to use cooling devices such as Exhaust fans or Air Conditioners.

Smart Home automation is increasingly gaining popularity among current applications of Internet of Things (IoT) due to the convenience and facilities it provides to the home owners. Sensors are employed within the home appliances via wireless connectivity to be accessible remotely by home owners to operate these devices. With the exponential increase of smart home IoT devices in the marketplace such as door locks, light bulbs, power switches etc, numerous security concerns are arising due to limited storage and processing power of such devices, making these devices vulnerable to several attacks. Due to this reason, security implementations in the deployment of these devices has gained popularity among researchers as a critical research area. Moreover, the adoption of traditional security schemes has failed to address the unique security concerns associated with these devices. Blockchain, a decentralised database based on cryptographic techniques, is gaining enormous attention to assure security of IoT systems. The blockchain framework within an IoT system is a fascinating substitute to the traditional centralised models, which has some significant concerns in fulfilling the demand of smart homes security. In this article, we aim to examine the security of smart homes by instigating the adoption of blockchain and exploring some of the currently proposed smart home architectures using blockchain technology. To present our findings, we describe a simple secure smart home framework based on a refined version of blockchain called Consortium blockchain. We highlight the limitations and opportunities of adopting such an architecture. We further evaluate our model and conclude with the results by designing an experimental testbed using a few household IoT devices commonly available in the marketplace.