Embedded System Research Articles

A Novel Approach for Early Breast Cancer Detection based on Embedded Micro-Bio Heat Ultrasensitive Sensors: IoT Technology

Several diagnostic methods are currently used for providing data about the existence of breast cancer. The aim of this work is to propose a new method of diagnosis using thermography, based on the use of an embedded ultrasensitive micro-bio-heat sensors 3x3 array modelled by Altium, realized as an embedded system for the early detection of breast cancer. This system benefits from IoT technology which offers the possibility of communicating via Smartphone with a server for processing the data subsequently consulted by the treating physician, the patient and/ or a possible medical monitoring centre.

Real-time fault diagnosis system for electrical panel using embedded systems

Regular maintenance inspections are necessary for industrial electrical panels. The project incorporated an IoT system that leverages embedded systems to streamline and expedite this process. This implementation rendered the electrical panel in real-time, enabling more efficient maintenance procedures. The project utilized an Odroid embedded system board as the central system controller and designed an electronic circuit for data communication. Software design was completed using the C++ programming language, and Dataplicity Cloud Commander was used as an interface for accessing the electrical panel online via phones or computers. This design allows technicians to continuously monitor the panel’s protective relays, temperatures, and power supply voltages. The system underwent testing for various operating conditions and potential errors, resulting in a low-cost intelligent system that significantly reduces troubleshooting time for electrical panels by providing essential information quickly and easily.

Teaching embedded control system design of electromechanical devices using a lab-scale smart farming system

This article presents the design and monitoring of a lab-scale smart farming system through the integration of control and app designs that can be used for teaching embedded control application to electromechanical systems. A combination of sensors and actuators is used to develop an Arduino-based embedded feedback control system that could be implemented in a smart farming environment. Specifically, we look at controlling electromechanical devices to actuate the fan and water pump to provide the optimal temperature and moisture, respectively, to enhance plant growth in a smart farming setting. The effectiveness of the feedback control is tested by conducting a plant growth experiment. Using garden cress ( Lepidium sativum) as a case study, the plant grown in the controlled temperature and moisture settings shows substantially healthier growth compared to the one grown in the non-controlled environment. In addition, an app is designed and developed to transform the Arduino data stream from the sensors into valuable insights that could help the users to monitor and improve the overall crop health. The developed system in this paper enables students to learn integral skills from interdisciplinary engineering fields (e.g. systems, control, mechanical and computer) to solve an agricultural problem.

Applications of Embedded Systems in Medicine: Challenges and Future Trends

With the advancement of science and technology, embedded systems play an essential role in the medical field. They have become an integral part of medical equipment, enabling functions such as patient monitoring, diagnosis, and treatment. Embedded systems also improve the work efficiency and accuracy of medical staff. This paper summarizes and reviews the recent applications of embedded systems in medicine. Firstly, it discusses the advantages and challenges of embedded system applications in medicine. Then, it discusses the development prospects and directions of embedded systems in the medical field. Through analysis, it is found that the application of embedded systems in the medical field is developing towards intelligence, miniaturization, and personalization. In the future, embedded systems will become more significant in the medical field, providing convenience to patients with different medical needs and bringing more innovations and changes to the medical field.

Intelligent Handling Robot Based on Mecanum Wheel

This paper proposes an intelligent handling robot based on Mecanum wheel, which can be set to recognize and handle the goods in the inherently known site, and can be adapted to the logistics sorting and other scenarios. The robot to NXP's RT1064 microcontroller as the main control, the use of RT-Thread embedded system development, the Mecanum wheel as a framework, the independent construction of the car model structure, equipped with a Raspberry Pi and robotic arm structure, as well as the MT9V032 total drilling wind camera to build the hardware circuit. The robot can recognize the position of the handling object through the off-site camera, then the upper computer for path planning, through the PID parameter adjustment and vehicle attitude solving to control the movement of the handling robot, to reach the handling location and then by the total drilling wind camera for detailed position adjustment. The robot equipped with Openart mini can recognize the type of objects to be handled, pick them up and place them on the pallet through robotic arm attitude solving, and correct the deviation through remote manual control, and finally realize fast omni-directional handling through Mecanum wheels to deliver the objects to the end point.

SReResNet: A stage recursive residual network for suppressing semantic redundancy during feature extraction

Feature extraction neural networks are essential components of computer vision systems. As the most famous one, ResNet has been widely used in engineering. Although many modern networks have outperformed ResNet, the costly pretraining and hyper-parameter optimization processes prevent their application of them in industrial computer vision systems. To avoid these costly processes, a Stage Recursive Residual Network (SReResNet) is proposed, which merely adjusts the forward propagation pipeline without changing the parameter architecture of ResNet. Thus, it can inherit existing model parameters of ResNet and replace ResNet through simple fine-tuning. Moreover, SReResNet is the first network to improve accuracy by utilizing the semantic trend during feature extraction instead of well-designed modules. It models a series of cascaded modules as a semantic unit and feeds the high-level feature maps back to the low-level modules for further semantic redundancy suppression through one feedback connection within the unit based on the semantic trend and the mechanism of looking and thinking twice. For object detection tasks, a Stage Recursive Feature Pyramid Network (SReFPN) is also proposed to rethink and suppress semantic redundancy further. Experiments demonstrate that SReResNet outperforms its counterparts in object detection and image classification tasks. On the MS COCO 2017 dataset, SReResNet outperforms ResNet with 1.2 Box AP improvement, and SReResNet with SReFPN further achieves 2.5 Box AP improvement without bells and whistles. On the CIFAR-100 dataset, SReResNet outperforms its counterparts, such as ResNet, DenseNet, and ConvNeXt, with at least 2.33% top-1 acc improvement. The code is available at https://github.com/unbelieboomboom/SReResNet.

Hybrid SOC and SOH estimation method with improved noise immunity and computational efficiency in hybrid railroad propulsion system

The fuel-cell and lithium-ion battery (LIB) hybrid railroad propulsion system (HRPS) technologies are actively being researched to reduce greenhouse gas (GHG) emissions. It is becoming increasingly important to accurately estimate the state-of-charge (SOC) and state-of-health (SOH) to enhance the safety and efficiency of the system. For a real-time embedded system, computational efficiency and the effect of noise are crucial challenges for SOC and SOH estimation. To overcome these problems, a method is required that improves computational efficiency and minimizes the effect of noise.To reduce computational cost and the effect of noise, the adaptive extended Kalman filter (AEKF) and specific partial capacity with single forgetting factor recursive least square (SPCSFFRLS) methods were combined to estimate SOC and SOH. In AEKF-based SOC estimation, the state and measurement noise covariance matrices were corrected by the mean square error (MSE) of the difference between measured voltage and modeled voltage. Thus, AEKF is effective in conquering measurement noise. The battery capacity was obtained by the SPCSFFRLS using station-to-station operation data. To calculate true specific partial capacity (SPC) without noise, the partial capacity method and SFFRLS were used to update the battery capacity. However, as the current accumulation period becomes longer, an accumulation error due to measurement noise may occur. Therefore, a specific period was analyzed to reduce accumulation error caused by measurement noise in the HRPS load profile. Finally, SOC and SOH co-estimation was demonstrated for a single cell and a parallel-connected battery pack (1S18P). As a result, it was found that the proposed method can reduce the computational time by about 33.0 % average while keeping the SOC and SOH estimation accuracy compared to DEKF. Thus, the proposed method is suitable for real-time embedded system such as HRPS to estimate SOC and SOH.

Online Outdoor Terrain Classification Algorithm for Wheeled Mobile Robots Equipped with Inertial and Magnetic Sensors

Terrain classification provides valuable information for both control and navigation algorithms of wheeled mobile robots. In this paper, a novel online outdoor terrain classification algorithm is proposed for wheeled mobile robots. The algorithm is based on only time-domain features with both low computational and low memory requirements, which are extracted from the inertial and magnetic sensor signals. Multilayer perceptron (MLP) neural networks are applied as classifiers. The algorithm is tested on a measurement database collected using a prototype measurement system for various outdoor terrain types. Different datasets were constructed based on various setups of processing window sizes, used sensor types, and robot speeds. To examine the possibilities of the three applied sensor types in the application, the features extracted from the measurement data of the different sensors were tested alone, in pairs and fused together. The algorithm is suitable to operate online on the embedded system of the mobile robot. The achieved results show that using the applied time-domain feature set the highest classification efficiencies on unknown data can be above 98%. It is also shown that the gyroscope provides higher classification rates than the widely used accelerometer. The magnetic sensor alone cannot be effectively used but fusing the data of this sensor with the data of the inertial sensors can improve the performance.

A three‐phase three‐level converter with shared flying capacitors

AbstractThree‐phase multilevel converters use a high number of components which increases their size. As specific weight reduction is a work of interest for embedded system applications, this paper presents a topology aiming to optimize volume and mass for a three‐level three‐phase inverter named 3L Hybrid. This three‐phase topology reduces the stored energy by two‐thirds compared with three‐level flying capacitor (FC) while keeping the same characteristics. The ‘phase‐sharing’ principle allowing the reduction number of FCs is presented and the three‐phase 3L Hybrid working will be described. An original closed‐loop control method has been developed for this structure. The control solution is generic and can be applied on other common flying capacitor (CFC) topologies. Finally, experimental results are presented and comparative data with well‐known multilevel topologies are analysed.

Flood Monitoring and Early Warning Systems – An IoT Based Perspective

One of the most frequently occurring calamities around the world is the flood. For flood prone areas or countries, an essential part of their governance is flood management. The necessity to continuously review and analyse the adverse or ambient environmental conditions in real-time demands developing a monitoring system so that floods could be detected beforehand. This paper discusses different Internet of Things (IoT) based techniques and applications implemented for efficient flood monitoring and an early warning system and it is observed that in future, the combination of IoT and Synthetic Aperture Radar (SAR) data may be helpful to develop robust and secure flood monitoring and early warning system that provides effective and efficient mapping during natural disasters. The emerging technology in the discipline of computing is IoT, an embedded system that enables devices to gather real-time data to further store it in the computational devices using Wireless Sensor Networks (WSN) for further processing. The IoT based projects that can help collect data from sensors are an added advantage for researchers to explore in providing better services to people. These systems can be integrated with cloud computing and analyzing platforms. Researchers recently have focussed on mathematical modeling based flood prediction schemes rather than physical parametric based flood prediction. The new methodologies explore the algorithmic approaches. There have been many systems proposed based on analog technology to web-based and now using mobile applications. Further, alert systems have been designed using web-based applications that gather processed data by Arduino Uno Microcontroller which is received from ultrasonic and rain sensors. Additionally, the machine learning based embedded systems can measure different atmospheric conditions such as temperature, moisture, and rains to forecast floods by analyzing varying trends in climatic changes.

Research on Road Sign Detection and Visual Depth Perception Technology for Mobile Robots

To accomplish the task of detecting and avoiding road signs by mobile robots for autonomous running, in this paper, we propose a method of road sign detection and visual depth perception based on improved Yolov5 and improved centroid depth value filtering. First, the Yolov5 model has a large number of parameters, a large computational volume, and a large model size, which is difficult to deploy to the CPU side (industrial control computer) of the robot mobile platform. To solve this problem, the study proposes a lightweight Yolov5-SC3FB model. Compared with the original Yolov5n model, the Yolov5-SC3FB model only loses lower detection accuracy, the parameter volume is reduced to 0.19 M, the computational volume is reduced to 0.5 GFLOPS, and the model size is only 0.72 MB, making it easy to deploy on mobile robot platforms. Secondly, the obtained depth value of the center point of the bounding box is 0 due to the influence of noise. To solve this problem, we proposed an improved filtering method for the depth value of the center point in the study, and the relative error of its depth measurement is only 2%. Finally, the improved Yolov5-SC3FB model is fused with the improved filtering method for acquiring centroid depth values and the fused algorithm is deployed to the mobile robot platform. We verified the effectiveness of this fusion algorithm for the detection and avoidance of road signs of the robot. Thus, it can enable the mobile robot to correctly perceive the environment and achieve autonomous running.

An Ultra-low-power Embedded AI Fire Detection and Crowd Counting System for Indoor Areas

Fire incidents in residential and industrial areas are often the cause of human casualties and property damage. Although there are existing systems that detect fire and monitor the presence of people in indoor areas, research on their implementation in embedded platforms is limited. This article introduces an ultra-low-power embedded system for fire detection and crowd counting using efficient deep learning methods. For the prediction of fire occurrences, environmental and gas sensor along with multilayer perceptron nodes are used. For crowd counting, a custom lightweight version of YOLOv5 is introduced, using an architecture based on ShuffleNetV2, resulting in a model with low memory requirements, high accuracy predictions, and fast inference on an embedded platform. The accuracy, power consumption, and memory requirements of the proposed system are evaluated using public datasets and datasets acquired by the environmental and image sensors, and its performance is compared to that of existing approaches.

ROS-based Controller for a Two-Wheeled Self-Balancing Robot

In this article, a controller based on a Robot Operating System (ROS) for a two-wheeled self-balancing robot is designed. The proposed ROS architecture is open, allowing the integration of different sensors, actuators, and processing units. The low-cost robot was designed for educational purposes. It used an ESP32 microcontroller as the central unit, an MPU6050 Inertial Measurement Unit sensor, DC motors with encoders, and an L298N integrated circuit as a power stage. The mathematical model is analyzed through Newton-Euler and linearized around an equilibrium point. The control objective is to self-balance the robot to the vertical axis in the presence of disturbances. The proposed control is based on a bounded saturation, which is lightweight and easy to implement in embedded systems with low computational resources. Experimental results are performed in real-time under regulation, conditions far from the equilibrium point, and rejection of external disturbances. The results show a good performance, thus validating the mechanical design, the embedded system, and the control scheme. The proposed ROS architecture allows the incorporation of different modules, such as mapping, autonomous navigation, and manipulation, which contribute to studying robotics, control, and embedded systems.

FiMec tremor stabilization spoon: design and active stabilization control of two DoF robotic eating devices for hand tremor patients.

This article is about vibration-damping robotic eating devices designed for use by people who have difficulty in eating due to hand tremors due to neuromuscular system disorder. The robotic eating device has two degrees of freedom (DoF). It contains an active controller structure to absorb vibrations in the y- and z-directions. In the handle part of the robotic eating device, there are two DC motors placed on the y- and z-axis, a three-axis IMU inertia sensor, an embedded system board, and a power unit. To absorb the vibration measured from the IMU sensor, the position control of the two motors to which the spoon is connected is provided by PID controllers. The part of the spoon (the pit surface) where the food is placed is tried to be kept constant. To test the vibration-damping performance of the control method, the dynamic model of the spoon along the eating kinematic trajectory was simulated in the SimMechanics environment using vibration data from ten tremor patients. The results show that the stabilization method can absorb the vibration in the hand of the person in the range of 84-99.409% and successfully provide the stabilization of the spoon tip. This damping rate is promising for providing a healthy diet for hand tremor patients.

3D LIDAR SEGMENTATION BASED ON EUCLIDEAN CLUSTERING FOR EMBEDDED SYSTEM

This paper presents a method for 3D LiDAR segmentation based on Euclidean clustering specifically designed for embedded systems. LiDAR sensors are widely used for perception tasks in autonomous vehicles, robotics, and other applications, providing dense point cloud data of the environment. Segmentation of the point cloud into meaningful objects is essential for understanding the surroundings and making informed decisions. Euclidean clustering is an effective approach for grouping points based on spatial proximity, enabling object segmentation. However, implementing such algorithms on embedded systems poses challenges due to limited computational resources. In this work, we propose an optimized implementation of the Euclidean clustering algorithm tailored for embedded systems to achieve real-time performance on the embedded system. The proposed approach involves acquiring raw point cloud data from the LiDAR sensor and preprocessing it to reduce noise and size. Adaptive Euclidean clustering is then applied to group points into clusters based on their spatial proximity. Extracted features such as centroids and bounding boxes are utilized for object classification and segmentation. Post-processing steps refine the segmentation results, improving accuracy and removing spurious clusters

Drowsy Eyes and Face Mask Detection for Car Drivers using the Embedded System

Efforts to prevent the spread of the COVID-19 virus have underscored the critical importance of mask-wearing as a preventive measure. Concurrently, road traffic accidents, often resulting from human error, have emerged as a significant contributor to global mortality rates. This study endeavors to address these pressing issues by employing advanced Deep Learning techniques to detect mask usage and identify drowsy eyes, thus contributing to the prevention of COVID-19 and accidents due to driver fatigue. To achieve this objective, an embedded system was developed, utilizing the integration of hardware and software components. The system effectively utilizes MobileNetV2 for face mask detection and employs HOG and SVM algorithms for drowsy eye detection. By seamlessly integrating these detection systems into a single embedded device, the simultaneous detection of both mask usage and drowsy eyes is made possible. The results demonstrates a commendable accuracy rate of 80% for face mask detection and 75% for drowsy eye detection. Furthermore, the mask detection component exhibits a remarkable training accuracy of 99%, while the drowsy eye detection component demonstrates an 80% training accuracy, affirming the system's efficacy in precisely identifying masks and drowsy eyes. The proposed embedded system offers potential applications in enhancing road safety. Its capability to effectively detect drowsy eyes and mask usage in car drivers contributes significantly to preventing accidents due to driver fatigue. Additionally, it plays a vital role in mitigating COVID-19 transmission by promoting widespread mask-wearing among individuals. This study exemplifies the potential of integrating Deep Learning methodologies with embedded systems, thus paving the way for future research and development in the realm of driver safety and virus prevention.

A monolithic light scattering particle CMOS sensor for integrated optics application

Monolithic CMOS sensor Particulate matter (PM) pollution in the air seriously endangers human health. The concentration of PM becomes an important indicator of the air quality. Traditional PM detector is generally based on MCU system, in which an embedded algorithm is used for counting the number of the scattered pulses. In the paper, a monolithic light scattering particle CMOS sensor is proposed and designed. The novel PM sensor IC, which combines the photometer method with CMOS circuit, can replace the traditional embedded system, with the benefits of high speed, high precision while low cost. The designed sensor shows the detection speed less than 2 μs, the detection error less than 0.01 mg/m3, and the overall layout area less than 1 mm2. The novel particle CMOS sensor proposed in this paper shows rapid and high-precision detection, while also reducing the manufacturing cost. It can be applied widely in the particle sensing scenarios, including the air pollution and water quality checking, etc..

Security assessment of in-vehicle communication protocols

As the automotive industry has progressed in recent years, in-vehicle networks have also changed. In contrast to traditional vehicles, modern vehicles communicate through connecting interfaces, Electronic Control Units (ECUs), sensors, and gateways over different communication channels called buses. While LIN, CAN, FlexRay, MOST, and Automotive Ethernet protocols facilitate vehicle communication, they expose vehicles to various security risks. Electronics in vehicles must be advanced to improve passenger safety, comfort, and vehicle performance. However, security issues must also be addressed to ensure vehicle resilience. This paper presents an in-depth and comprehensive analysis of the security issues associated with in-vehicle network protocols. We examine the feasibility, severity, and potential consequences of these security issues, which can be used to better understand the risks and vulnerabilities within in-vehicle networks. Furthermore, the paper examines existing solutions that aim to mitigate the identified attacks, including intrusion detection systems (IDSs), firewalls, authentication, and schedulers, while evaluating their effectiveness and drawbacks. Hence, we enable automotive security engineers to comprehend the current state of security measures used to protect in-vehicle networks and identify areas where further research is needed to develop more effective and tailored security solutions. Finally, the paper provides recommendations for future research and the development of security solutions that researchers can use to guide their research efforts for in-vehicle network security.

An Embedded System Based on Raspberry Pi for Effective Electrocardiogram Monitoring

In recent years, there has been a growing demand for affordable and user-friendly medical diagnostic devices due to the rise in global diseases. This study focuses on the development of an embedded system based on Raspberry Pi that enables faster and more efficient monitoring of electrocardiogram (ECG). The incorporation of Raspberry Pi allows for both wireless and wired interfaces, facilitating the creation of an ECG diagnostic embedded system capable of real-time detection and immediate response to any abnormalities in heart functionality. The system presented in this research encompasses a comprehensive electronic circuit comprising analog and digital components to measure and display the ECG signal. Within the analog section, the circuit performs essential signal conditioning tasks, such as signal amplification and noise filtering, ensuring a clean signal within the desired frequency range. The entire system is powered using a power bank. The digital segment incorporates an analog-to-digital converter necessary for converting the received analog signal into a digital format compatible with Raspberry Pi. A graphical liquid-crystal display is utilized to display the measured signal. The device successfully measures ECG signals at various heart rates, capturing all crucial peaks that can be used as indicators of an individual’s health condition. By comparing the signals obtained from healthy individuals with those exhibiting heart arrhythmias, valuable insights can be gained regarding their health status. The proposed system aims to be portable, cost-effective, and user-friendly in different environments.

A tractor hydraulic assisted embedded microprocessor-based penetrometer for soil compaction measurement

Measurement of soil compaction in terms of penetration resistance using a hand penetrometer requires more time and effort. A tractor hydraulic-assisted embedded microprocessor-based penetrometer was developed to make the results easy to monitor and decrease human error. The developed penetrometer consisted of a driving system for pushing the penetrometer probe into the soil at a desired speed, a sensor unit (load cell and ultrasonic sensor) for measuring the force required to push the probe and the depth of penetration, and a data logging system. The embedded system of the penetrometer comprised various components: Arduino Uno microcontroller, I2C, WiFi module, and HX711 amplifier. The system was able to measure a soil penetration resistance of 5000 kPa up to a depth of 500 mm. A WiFi module helped to transmit the data to an Android mobile application. The acquired data also include date/time parameters along with latitude and longitude locations picked up from the GPS of the mobile. Its performance was assessed based on sensor characteristics such as accuracy, sensitivity, non-linearity, and non-repeatability. The average and maximum absolute variations in the cone index values measured with developed system and a hand-held digital penetrometer were observed to be 22.50% and 27.88%, respectively.