Embedded System Research Articles

Air Quality Analysis through IoT Device and Risk Prediction of Asthma Attack using ML Techniques

Background: The air quality of any area depends upon the various PMs (particulate matter) and hazardous gases present in the air. Low-cost PM sensors and gas sensors are present in different target places to monitor the air quality, read the environmental data, and transmit it to local servers through the IoT device. The low-cost sensor is not reliable due to its low sensing capacity; therefore, the read data is calibrated with the meteorological data presented by the nearby meteorological Centre of that particular area. The calibrated reading data sent to the server could be analyzed through some Machine Learning [ML] models. The ML models help to predict the risk of asthma in a particular area. The risk of asthma is directly related to the air quality of the surroundings. It is observed that the air quality of the industrial area is much worse than the non-industrial belt. Air quality monitoring of industrial areas is always a challenging task due to the ununiformed pollution in some segregated places around the industry, emitting pollutants mostly from chimneys. The air quality of any area depends upon the PM (PM), i.e., PM2.5 and PM10.0, as well as the gasses like NO2(Nitrogen Dioxide), NH3 (Ammonia), SO2(Sulfur dioxide), CO(Carbon monoxide), O3(Ozone) and Benzene. These are the most hazardous gases generally emitted by common heavy industries like iron and steel. In this article, the researchers considered the industrial belt of the Asansol- Durgapur region of West Bengal, India, and predicted the risk of asthma attacks for the test dataset. The experiment was carried out on 10 different supervised machine learning [SML] models as well as semi-supervised machine learning (SSML) models. The SML models have been further refined through hyper-parameter tuning, and better results have been obtained in the case of some ML models. The result has been compared with the existing literature considering the same external environment from where the meteorological data was collected, and similar ML models have been used. The research outperformed the existing literature, which is depicted in the result and analysis section of the article. Methods: The study evaluated ML models, both supervised and semi-supervised, to assess pollution levels. Relevant features were selected while less relevant ones were discarded. Accuracy levels of different ML algorithms werecompared in the results. The most effective model for an IoT system was chosen to maximize accuracy. In semi-supervised learning, feature selection followed supervised learning, but testing was akin to unsupervised learning. Results were compared with supervised learning data, enhancing reliability. Results: The result employing various classifiers werepresented across tables after the independent parameter Ozone was removed. Following the output of several classifiers, the results were verified using the k-fold validation method, with k being set to 5 or 10, accordingly. Tables display the best outcome, which is indicated in bold characters. method: In this research work the researcher considered 9 different ML models and used them as supervised as well as semi supervised model to determine the pollution level of the certain area. In this research work the researcher also selected the most relevant features and discarded the less relevant features. In case of SML algorithm, the accuracy level of the different ML algorithm has been determined and depicted in the result analysis section. The most effective ML model has been chosen for the proposed embedded system so that accuracy could be achieved at most. In case of semi supervised algorithm the feature selection is done as per the supervised algorithm. In this case the training is done same as the SML algorithm, but the testing phase is done like unsupervised machine learning algorithm where the decision parameter is predicted and ultimately matched with the previously achieved data of SML algorithm. The reliability of this approach is much more effective than simple SML algorithm. Conclusion: This study focused on predicting asthma risk in the Asansol-Durgapur industrial belt, India, using low-cost PM and gas sensors. Data calibration with meteorological inputs enhanced accuracy. ML models predicted risk and were refined through hyper-parameter tuning. Comparative analysis showed superior performance, emphasizing the importance of precise air quality monitoring. While offering a robust framework for future research, the study’s limitation lies in its area-specific dataset.

Adaptive Detection in Real-Time Gait Analysis through the Dynamic Gait Event Identifier.

The Dynamic Gait Event Identifier (DGEI) introduces a pioneering approach for real-time gait event detection that seamlessly aligns with the needs of embedded system design and optimization. DGEI creates a new standard for gait analysis by combining software and hardware co-design with real-time data analysis, using a combination of first-order difference functions and sliding window techniques. The method is specifically designed to accurately separate and analyze key gait events such as heel strike (HS), toe-off (TO), walking start (WS), and walking pause (WP) from a continuous stream of inertial measurement unit (IMU) signals. The core innovation of DGEI is the application of its dynamic feature extraction strategies, including first-order differential integration with positive/negative windows, weighted sleep time analysis, and adaptive thresholding, which together improve its accuracy in gait segmentation. The experimental results show that the accuracy rate of HS event detection is 97.82%, and the accuracy rate of TO event detection is 99.03%, which is suitable for embedded systems. Validation on a comprehensive dataset of 1550 gait instances shows that DGEI achieves near-perfect alignment with human annotations, with a difference of less than one frame in pulse onset times in 99.2% of the cases.

Design of an Embedded Test Bench for Organic Photovoltaic Module Testing

In this article, a multipurpose embedded system for testing organic photovoltaic modules is presented. It is designed to include all the features for real-time monitoring, data acquisition, and power conversion based on a Ćuk converter, providing useful data for scientific investigation of the outdoor operation of organic photovoltaic modules. The embedded system allows both the scan of the I–V curve and the continuous operation of the organic photovoltaic module, such as at its maximum power. Voltage and current at the terminals of the organic photovoltaic module under test and up to four temperatures are continuously measured and stored on a Secure Digital card. The communication interface allows the embedded system to connect with other instruments, such as irradiance sensors, with digital serial output. The embedded system is designed both for laboratory and in-the-field use: it can be powered either by the AC electrical grid or a battery, which can also operate as a backup battery. Galvanic isolation divides the embedded system into the field-side and the logic-side functional sections, providing improved noise immunity and safe operation. The main power distribution system within the embedded system is a +9 V bus; ultra-low-noise linear low dropout regulators provide the +3.3 V and +5 V regulated voltages to supply the analog and digital circuits within the logic-side section, and a flyback converter supplies the field-side section of the board. The proposed embedded solution is validated using an experimental setup built at SolarTechLab, Politecnico di Milano. The experimental results report the feasibility of the proposed embedded system.

Developing a custom communication protocol for UAVs: Ground control station and architecture design

Robots have become a prominent research topic across various application domains in recent decades. Additionally, unmanned aerial vehicles (UAVs) are extensively used in both the military and commercial sectors, substantially reducing transaction costs and enhancing safety. Researchers have addressed secure control and communication protocols between software, firmware, and hardware components. This study focuses on the design of three critical elements: the hardware architecture, the software ground control station (GCS), and the firmware tasks within the UAV embedded system. These components are interconnected via an enhanced MAVLink protocol (EMP). Furthermore, various sensors are integrated into the UAV's peripheral devices. We discuss flight control (FC) approaches, such as proportional-integral-derivative (PID) control and the Kalman filter (KF), detailing the process of the hovering algorithm. Additionally, we explain how access is messaged and how message commands are implemented at the protocol layer. We propose a large-scale UAV system architecture suitable for commercial and military applications, supported by a real-life scenario. Experimental results demonstrate the effectiveness and efficiency of the UAV in outdoor activities. Our findings confirm that the proposed UAV architecture is a robust and efficient system in practical applications.

Smart ArM: a customizable and versatile robotic arm prosthesis platform for Cybathlon and research

BackgroundIn the last decade, notable progress in mechatronics paved the way for a new generation of arm prostheses, expanding motor capabilities thanks to their multiple active joints. Yet, the design of control schemes for these advanced devices still poses a challenge, especially with the limited availability of command signals for higher levels of arm impairment. When addressing this challenge, current commercial devices lack versatility and customizing options to be employed as test-beds for developing novel control schemes. As a consequence, researchers resort to using lab-specific experimental apparatuses on which to deploy their innovations, such as virtual reality setups or mock prosthetic devices worn by unimpaired participants.MethodsTo meet this need for a test-bed, we developed the Smart Arm platform, a human-like, multi-articulated robotic arm that can be worn as a trans-humeral arm prosthesis. The design process followed three principles: provide a reprogrammable embedded system allowing in-depth customization of control schemes, favor easy-to-buy parts rather than custom-made components, and guarantee compatibility with industrial standards in prosthetics.ResultsThe Smart ArM platform includes motorized elbow and wrist joints while being compatible with commercial prosthetic hands. Its software and electronic architecture can be easily adapted to build devices with a wide variety of sensors and actuators. This platform was employed in several experiments studying arm prosthesis control and sensory feedback. We also report our participation in Cybathlon, where our pilot with forearm agenesia successfully drives the Smart Arm prosthesis to perform activities of daily living requiring both strength and dexterity.ConclusionThese application scenarios illustrate the versatility and adaptability of the proposed platform, for research purposes as well as outside the lab. The Smart Arm platform offers a test-bed for experimenting with prosthetic control laws and command signals, suitable for running tests in lifelike settings where impaired participants wear it as a prosthetic device. In this way, we aim at bridging a critical gap in the field of upper limb prosthetics: the need for realistic, ecological test conditions to assess the actual benefit of a technological innovation for the end-users.

Embedded system controlled switching system for reconfigurable antennas

AbstractA multiband and reconfigurable antenna is the one of the main aspects in the communication system which has an extremely important role. An embedded based solution for the fast operation of a reconfigurable antenna to work in a specific operating frequency band is the solution of future technologies. Embedded based solution increases the speed by avoiding the dependence on manual control of the RF switching elements. In this paper, the proposed embedded system‐controlled switching mechanism operate the reconfigurable antenna in any one of the operating bands by a single command to be send from a remote location. The embedded system receives the command and configure a set of PIN diodes as per the modes. The proposed solution consists an Arduino UNO board as motherboard, SIM900A as GSM module to receive the command from user, biasing circuit to control the operation of PIN diode, PIN diode to reconfigure the patch pattern and a user mobile for sending AT command.The proposed model is successful and tested on proteus software 8.0. The model is also deployed on the newly designed multiple‐input‐multiple‐output antenna where the result shows the successful operation to achieve multiband frequency operation.

Robust identification of users by convolutional neural network in MATLAB and Raspberry Pi

The following article presents the development of an algorithm embedded in a Raspberry Pi 3B board, where a user identification was made, using the convolutional neural network (CNN) for 5 predefined users, with the option of loading remotely a new network for a new user. Comparatively, the same application was programmed in MATLAB programming software to evaluate the results and identify the advantages between them. Networks were trained for 5 different users, using the Caffe library on the Raspberry Pi, and the MATLAB neural network package on the computer. Where it was found that the training made by Caffe on an embedded system is much slower and less efficient than the ones performed in MATLAB, obtaining less than 55% accuracy with Caffe networks and more than 90% with MATLAB networks, training with the same number of samples, the same architecture, and the same database. Finally, the accuracy obtained through confusion matrix is over 88% in each case of users identification.

Internet of things and radio frequency identification based embedded system to reduce shopping time in supermarkets

Doing daily shopping in a Peruvian supermarket means a large investment of time for many people, usually due to inaccurate and faulty scanning of products by barcodes at supermarket checkout counters. For this reason, an embedded system based on internet of things (IoT) and radio frequency identification (RFID) is designed to reduce shopping time in a supermarket. The system uses an ESP32 development board with embedded hardware specialized in IoT projects and firmware development based on C language and real-time operating systems (FreeRTOS) through espressif’s IoT development framework (ESP-IDF). RFID tags were used to scan the products and IoT with message queuing telemetry transport (MQTT) communication protocol are implemented to a local database in real time. The system achieves a significant reduction in terms of scanning time compared to self-service checkouts using barcodes, which allows to statistically analyze the reduced time per quantity of products and the linear trend of the 2 samples.

An FPGA-based parallel deconvolution application envisaging high-energy calorimeters operating under severe pile-up conditions

Signal pileup may arise when the system response time is larger than what would be required by the event rate in the application. In this condition, information superposition occurs, and signal estimation is deteriorated due to such information distortion. For online applications, mitigating such a pileup distortion usually requires embedded digital signal processing, which often involves FPGA-based designs when high event rates are a concern. Recently, high-energy particle experiments started facing an unprecedented increase in pileup conditions, which demanded new approaches for signal estimation. This work proposes a dedicated parallel signal processing system embedded on an FPGA platform that implements different filtering techniques for online energy reconstruction in calorimeters (energy measurement) operating under severe pileup conditions. Each designed filter is an inverse approximation of the calorimeter readout channel's impulse response, performing deconvolution of the incoming signals. Such a deconvolution approach for energy estimation was recently introduced, and the proposed filters were implemented in this work. Evaluation of the digital system design was performed using a data set that considers a typical high-energy calorimeter front-end response, as high-luminosity particle collider experiments produce high pileup distortions in calorimeter signals when processing the subproducts of the collisions. As the Large Hadron Collider (LHC) provides the most demanding conditions in terms of signal pileup, the acquisition system of the proposed solution is synchronous with its collision rate (40 MHz). The results showed that the proposed implementation may be applied as a preprocessing (pileup reduction) step in calorimeter instrumentation and could reduce the signal pileup within the fast response time required by such experiments.

CNN-LSTM for MFCC-based Speech Recognition on Smart Mirrors for Edge Computing Command

Smart mirrors are conventional mirrors that are augmented with embedded system capabilities to provide comfort and sophistication for users, including introducing the speech command function. However, existing research still applies the Google Speech API, which utilizes the cloud and provides sub-optimal processing time. Our research aim is to design speech recognition using Mel-frequency cepstral coefficients (MFCC) and convolutional neural network–long short-term memory (CNN-LSTM) to be applied to smart mirror edge devices for optimum processing time. Our first step was to download a synthetic speech recognition dataset consisting of waveform audio files (WAVs) from Kaggle, which included the utterances “left,” “right,” “yes,” “no,” “on,” and “off. ” We then designed speech recognition by involving Fourier transformation and low-pass filtering. We benchmark MFCC with linear predictive coding (LPC) because both are feature extraction methods on speech datasets. Then, we benchmarked CNN-LSTM with LSTM, simple recurrent neural network (RNN), and gated recurrent unit (GRU). Finally, we designed a smart mirror system complete with GUI and functions. The test results show that CNN-LSTM performs better than the three other methods with accuracy, precision, recall, and an f1-score of 0.92. The speech command with the best precision is "no," with a value of 0.940. Meanwhile, the command with the best recall is "off," with a value of 0.963. On the other hand, the speech command with the worst precision and recall is "other," with a value of 0.839. The contribution of this research is a smart mirror whose speech commands are carried out on the edge device with CNN-LSTM.

Accurate and rapid reactive flow simulations using dynamic load balancing and sparse analytical Jacobian approach

The present study introduces a rapid and accurate customized solver on the OpenFOAM platform for large-scale industrial computations. Specifically, a sparse analytical Jacobian approach utilizing the SpeedCHEM library was implemented to enhance the efficiency of the ordinary differential equation solver. The dynamic load balancing code was used to distribute computational workloads uniformly across multiple processes. Optimization continued with open multi-processing to improve parallel computing efficiency and the local time stepping scheme to maximize individual cell time steps. The effectiveness and robustness of the customized solver were first validated using Sandia flames D–F as benchmarks. The results showed that the customized solver exhibited better strong scaling characteristics and led to a speed increase of up to 30 times for two-dimensional Sandia flame D calculations. The numerical predictions for temperature and species distribution closely matched the experimental trends, confirming the accuracy of the solver. Subsequently, a three-dimensional numerical study on a 10 kW ammonia co-combustion furnace was conducted, exploring the performance of the solver in large-scale reactive simulations. Results analysis indicated that the acceleration capability was reduced due to increased communication overhead between processors, achieving up to 7.06 times speed-up. However, as the size of the reaction mechanism increases, better acceleration capabilities can be demonstrated. The numerical predictions could closely replicate experimental trends, effectively predicting NO emission trends within the combustion furnace. This study offers one viable solution for rapid and accurate calculations in the OpenFOAM platform, which could be applied in the subsequent ammonia industrial combustion processes.

A Block-Wised and Sparsely-Connected ANN for the Prediction of Human Joint Moment Based on the Improved Hill Musculoskeletal Model

Human joint moment plays an important role in rehabilitation assessment and human-robot interaction, which cannot be measured directly but can only be predicted via indirect measurement by an artificial neural network (ANN). However, most existing ANN models for human joint moment prediction use fully-connected network which has complex structure and no inclusion of domain knowledge. Thus, this study introduced a novel block-wised and sparsely-connected ANN model (BSANN) for human joint moment prediction, which significantly reduced the computational and storage costs. In this BSANN model, by using an improved Hill musculoskeletal (HMS) model, a single-output fully-connected network was established as a block to take each electromyograph (EMG) signal for the prediction of the muscle moment, and all muscle moments were connected together as inputs to obtain the joint moment. Compared to the ANN, our BSANN model decreased 80.7% connections and keeps good prediction accuracy. It provides embedded portable systems a powerful tool to predict joint moment.

A real-time embedded system designed for NILM studies with a novel competitive decision process algorithm

A real-time embedded system designed for NILM studies with a novel competitive decision process algorithm

Comparison of KF-Based Vehicle Sideslip Estimation Logics with Increasing Complexity for a Passenger Car.

Nowadays, control is pervasive in vehicles, and a full and accurate knowledge of vehicle states is crucial to guarantee safety levels and support the development of Advanced Driver-Assistance Systems (ADASs). In this scenario, real-time monitoring of the vehicle sideslip angle becomes fundamental, and various virtual sensing techniques based on both vehicle dynamics models and data-driven methods are widely presented in the literature. Given the need for on-board embedded device solutions in autonomous vehicles, it is mandatory to find the correct balance between estimation accuracy and the computational burden required. This work mainly presents different physical KF-based methodologies and proposes both mathematical and graphical analysis to explore the effectiveness of these solutions, all employing equal tire and vehicle simplified models. For this purpose, results are compared with accurate sensor acquisition provided by the on-track campaign on passenger vehicles; moreover, to truthfully represent the possibility of using such virtual sensing techniques in real-world scenarios, the vehicle is also equipped with low-end sensors that provide information to all the employed observers.

Research on the Method and Application of Truck Type Recognition Based on Deep Learning from the Perspective of Unmanned Aerial Vehicles

A lightweight detection model for truck models based on improved YOLOv5s (MobileNetV3-YOLOv5s) is proposed to meet the requirement for real-time detection under the limited embedded device resources carried by drones. Firstly, we use MobileNetV3 to replace the backbone feature extraction network and use deep separable convolution to replace traditional convolution to reduce the model’s parameter count. Secondly, we use DIOU loss as the regression loss of the bounding box to enhance the convergence speed of the model and improve the ability to fit data. Finally, we use the K-means clustering method to reset the prior box. The experimental results show that the mAP value of the improved model is 89.6%, which is 0.2 percentage points lower than before, but the volume is only 3.98MB, which is about half of the original model. The detection speed is also significantly improved compared with before the improvement. Therefore, the lightweight model based on improved YOLOv5s improves detection speed and significantly reduces model volume while ensuring detection accuracy. It enables efficient, real-time recognition of truck models under complex road conditions on embedded devices.

The Artistic Impact of Computer-Aided Design on Building Façade in Nigeria

The visual landscape of our cities is ever- evolving, shaped by the architectural and embedded computer aided design marvels that adorn the urban fabric. Among the myriad elements that contribute to the allure of a building, the artistic impact on its facade holds a distinct significance. The facade, as the external face of a structure, acts as a canvas where artistry and design converge, manifesting an artistic expression that resonates with its surroundings and captivates the human eye. The interplay between art, computer and architecture has been a topic of fascination for centuries, as artists, computer aided design and architects have sought to transcend mere functionality and transform buildings into visually striking works of art. From ancient civilizations to modern metropolises, the artistic impact on building facades has been an enduring reflection of cultural heritage, societal values, and creative ingenuity. This journal aims to delve into the multifaceted aspects of artistic and computerization impact on building facades, exploring the diverse factors that shape this phenomenon and the profound influence it has on our built environment. By delving into the intricate interplay of aesthetics, functionality, and cultural context, we seek to unravel the power of artistic and automation expression in transforming buildings into iconic landmarks. Understanding the artistic and embedded system impact on building facades necessitates an exploration of various dimensions. Examining the influence of cultural context, as different regions and communities in Nigeria imbue their buildings with distinct artistic computerization styles, motifs, and themes rooted in their heritage and traditions. Additionally, architectural coupled with automation system style plays a pivotal role, as different design paradigms demand specific artistic elements and materials to achieve harmony and coherence between form and function. Furthermore, this journal will investigate the impact of the urban environment on artistic automation system expression. The surrounding cityscape, with its diverse architectural and computerization styles, urban planning principles, and contextual demands, serves as a backdrop against which building facades make their artistic and automation statements.

Design of visual symbol-aided system based on wireless network sensor and embedded system

Abstract In order to solve the problem of the low transmission rate of wireless network sensors (WNSs), this article proposes a WNS visual assistance system based on embedded systems. The system uses WNSs instead of traditional wired sensors, which can achieve synchronous transmission of multi-node data, enabling nodes to work together better, thereby improving the real-time and reliability of the entire system. This article conducts in-depth research on feature extraction algorithms and tests the visual assistance system in the experimental section. The results show that the recognition rate and stability of the visual symbol assistance system implemented using WNSs are higher than those of ordinary systems. In the satisfaction survey, it was found that 87 people were very satisfied with the visual symbol assistance system, accounting for 87%, while only 57 people were very satisfied with the traditional visual symbol assistance system, accounting for 57%. The experimental results show that the system output stability of the design method is good, and the response time and reliability are better.

Evaluating single event upsets in deep neural networks for semantic segmentation: An embedded system perspective

As the deployment of artificial intelligence (AI) algorithms at edge devices becomes increasingly prevalent, enhancing the robustness and reliability of autonomous AI-based perception and decision systems is becoming as relevant as precision and performance, especially in applications areas considered safety-critical such as autonomous driving and aerospace. This paper delves into the robustness assessment in embedded Deep Neural Networks (DNNs), particularly focusing on the impact of parameter perturbations produced by single event upsets (SEUs) on convolutional neural networks (CNN) for image semantic segmentation. By scrutinizing the layer-by-layer and bit-by-bit sensitivity of various encoder–decoder models to soft errors, this study thoroughly investigates the vulnerability of segmentation DNNs to SEUs and evaluates the consequences of techniques like model pruning and parameter quantization on the robustness of compressed models aimed at embedded implementations. The findings offer valuable insights into the mechanisms underlying SEU-induced failures that allow for evaluating the robustness of DNNs once trained in advance. Moreover, based on the collected data, we propose a set of practical lightweight error mitigation techniques with no memory or computational cost suitable for resource-constrained deployments. The code used to perform the fault injection (FI) campaign is available at https://github.com/jonGuti13/TensorFI2, while the code to implement proposed techniques is available at https://github.com/jonGuti13/parameterProtection.

Differentiable Geodesic Distance for Intrinsic Minimization on Triangle Meshes

Computing intrinsic distances on discrete surfaces is at the heart of many minimization problems in geometry processing and beyond. Solving these problems is extremely challenging as it demands the computation of on-surface distances along with their derivatives. We present a novel approach for intrinsic minimization of distance-based objectives defined on triangle meshes. Using a variational formulation of shortest-path geodesics, we compute first and second-order distance derivatives based on the implicit function theorem, thus opening the door to efficient Newton-type minimization solvers. We demonstrate our differentiable geodesic distance framework on a wide range of examples, including geodesic networks and membranes on surfaces of arbitrary genus, two-way coupling between hosting surface and embedded system, differentiable geodesic Voronoi diagrams, and efficient computation of Karcher means on complex shapes. Our analysis shows that second-order descent methods based on our differentiable geodesics outperform existing first-order and quasi-Newton methods by large margins.

Computer Vision Model Compression Techniques for Embedded Systems:A Survey

Deep neural networks have consistently represented the state of the art in most computer vision problems. In these scenarios, larger and more complex models have demonstrated superior performance to smaller architectures, especially when trained with plenty of representative data. With the recent adoption of Vision Transformer (ViT) based architectures and advanced Convolutional Neural Networks (CNNs), the total number of parameters of leading backbone architectures increased from 62M parameters in 2012 with AlexNet to 7B parameters in 2024 with AIM-7B. Consequently, deploying such deep architectures faces challenges in environments with processing and runtime constraints, particularly in embedded systems. This paper covers the main model compression techniques applied for computer vision tasks, enabling modern models to be used in embedded systems. We present the characteristics of compression subareas, compare different approaches, and discuss how to choose the best technique and expected variations when analyzing it on various embedded devices. We also share codes to assist researchers and new practitioners in overcoming initial implementation challenges for each subarea and present trends for Model Compression.