Globally, road transportation has become one of the most reliable means of moving goods and services from one place to the other. It has contributed immensely to the standard of living and modern civilization. However, this means of transportation is characterised by some issues which are poised to be harmful to the human population if not properly addressed. One of such issues is the presence of potholes, bumps, and other road anomalies. Unfortunately, the late identification of road anomalies (Speedbumps and Potholes) and the inability of drivers to detect and slow down while approaching such road anomalies has also been a big challenge faced by many nations. Therefore, there is a need for an automatic and intelligent approaches to be built into vehicles to mitigate the number of road accidents caused by these anomalies. In this work, the development of an intelligent road anomaly identification and manoeuvring system for autonomous vehicle is presented. The developed system focuses on the detection of road anomalies specifically speedbumps and potholes; and the regulation of the vehicular speed when these anomalies are detected. A modified Histogram Oriented Gradient (HOG) and Fuzzy Logic Control (FLC) have been proposed in this work. Furthermore, promising results have been obtained and presented which depicts the proposed HOG algorithm outweigh other techniques in the identification and detection of speedbumps and potholes. In addition, the developed FLC was able to regulate the speed of the vehicle in the presence of speedbumps as well as navigate the vehicle accordingly in the presence of potholes.

Read more…

This article focuses on risk analysis using a multi-criteria decision-making method. Due to many performed risk analyses for soft targets, we are constantly trying to find new methods for objective risk assessment. Many risk analyses are subjective, which is a problem when planning security measures and comparing results (different events, objects, places, etc.). In this text, we present our case study, which deals with the use of fuzzy TOPSIS. As a reference object, we have chosen one of the specific categories of soft targets – cultural events. The goal was to find the location most at risk of violent attacks on a selected cultural event – a music concert. We then established cooperation with three experts. The completed data in the risk analysis was then compared with practice. The selected fuzzy TOPSIS method was chosen as presumably more objective. Our hypothesis was confirmed. The results were objective and consistent with practical experiences.

Read more…

Recently, many problems have been caused by battery fires. The existing BMS(Battery Managment System) measured the voltage of each cell of the battery through the physical connection between the battery and the control module. However, if a battery with up to 1000 VDC becomes inoperable due to an external factor, the battery is damaged, and accordingly, a large current of the battery breaks the control unit of the BMS with 5 VDC to 24 VDC, putting the BMS inoperable. If the battery is operated when the bms is in trouble, it poses a risk of battery fire.Recently, as bms technology was announced with a wireless function, battery information could be easily transferred from the outside, so that convenience was maximized, but stability is still weak. This paper physically separated the battery and control module by measuring the battery voltage depending on the strength of the LED by connecting the battery and LED. and furthermore, the measurement error should be less than 1 mV even when the temperature changes. In addition, it was designed to operate at a low output level of 200 μW to 360 μW using the sub-threshold section of the LED.

Read more…

The present work shows the development of a data collection platform that allows the researcher to collect new video and voice data sets in Spanish. It also allows the application of a standardized personality test and stores this information to analyze the effectiveness of the automatic personality recognizers concerning the results of a standardized personality test of the same participant. Thus, it has elements to improve the evaluated models. These optimized models can then be integrated into intelligent learning environments to personalize and adapt the content presented to students based on their dominant personality traits. To evaluate the developed platform, an intervention was conducted to apply the standardized personality test and record videos of the participants. The data collected were also used to evaluate three machine learning models for automatic personality recognition.

Read more…

Despite many factors conducive to renewable energy investment in Qatar (e.g., the fact that the state is a major gas exporter whose long-term prosperity depends on economic diversification), there is very low uptake of solar panel adoption among home and business owners. Major challenges implicitly face the deployment of solar and other renewables in Qatar, this research explores possible challenges. The study was conducted in two phases: interviews to identify challenges and using the outcomes from the interviews to obtain a wider response. This study identifies the key major challenges facing the deployment of solar panels in Qatar, which are very useful for diverse stakeholders, policymakers, and future researchers.

Read more…

The Activity and Event Network (AEN) is a new dynamic knowledge graph that models different network entities and the relationships between them. The graph is generated by processing various network security logs, such as network packets, system logs, and intrusion detection alerts, which allows the graph to capture security-relevant activity and events in the network. In this paper, we show how the AEN graph model can be used for threat identification by introduc- ing an unsupervised ensemble detection mechanism composed of two detection schemes, one signature-based and one anomaly-based. The signature-based scheme employs an isomorphic subgraph matching algorithm to search for generic attack patterns, called attack fingerprints, in the AEN graph. As a proof of concept, we describe fingerprints for three main attack categories: scanning, denial of service, and password guessing. The anomaly-based scheme, in turn, works by extracting statistical features from the graph upon which anomaly scores, based on the bits of meta-rarity metric first proposed by Ferragut et al., are calculated. In total, 15 features are proposed. The performance of the proposed model was assessed using two intrusion detection datasets yielding very encouraging results.

Read more…

This study presents an “Omnidirectional multi-view image measurement system”, which can be used to provide multi-camera 3D reconstruction and multi-view image information. Its characteristic is that four cameras take images from multiple perspectives in the co-sphere framework. The C0 is the middle camera fixed as the geometric center point of measurement, and provides a front image. The other three cameras C1~C3 provide side images, and the co-circular spheres are separated by 120 degrees to extend the circle. The arc rod adjusts the multi-angle imaging. Place the multi-view camera in the arc track and move to the specified position in the sphere to position and capture images. By changing the angle between the cameras, the range of images captured by the cameras can be changed. If the multi-view images of four cameras C0, C1, C2 and C3 are captured at the same time, a stereo camera pair can be formed by any two cameras. The stereo camera pair C0-C1, C0-C2 and C0-C3 can be compiled by using the parallax principle of left and right images matching. Finally, through the demonstration and verification of camera calibration and 3D reconstruction, it can be used for all-round multi-view image measurement.

Read more…

This research presents a “Multi-camera for pilot’s cockpit measurement system”, which uses four multi-view images to eliminate the instrument and human body shielding and record the touched area. That could record the body reaction time (velocity and acceleration) and trajectory of the tested personnel. Real-time conversion of multi-view images corresponding to the 3D skeletal joint coordinate information of the human body, which measure the human-computer interaction human factors engineering integration of limb reaction time and trajectory measurement system. Finally, make prototypes, test and optimize, and achieve the research on the optimal cockpit touch area by conducting multi-view image simulation feasibility experiment framework and measurement process method. Using multiple depth-sensing cameras to perform low-cost, standardized automatic labeling of human skeleton joint dynamic capture.

Read more…

Power losses exist naturally and have to be cared for in the operation of electrical power systems. Many researchers have worked on various methods and approaches to reduce losses by incorporating distributed generators (DG), particularly from renewable sources. These studies are based on the maximum unit penetration of the DGs, which is rarely achieved, resulting in inaccurate calculations. This paper proposes an advanced solution for calculating power losses by incorporating an Extreme Learning Machine (ELM) method for forecasting the solar irradiation. The ELM algorithm was used to create a model for forecasting solar radiation in the Manokwari region and its surroundings. Daily solar radiation in the region has been predicted using the model. NASA’s 8016 data on temperature and solar irradiation were used to train the ELM model. With an MAE value of around 0.6392 and a training time of 4.4375 seconds, the test results demonstrate that the built model has good accuracy. The operation of a 1000 kWp solar power plant based on the ELM data forecasting can reduce the power loss of the existing distribution network around the location from 1.5095 kW/hour to 0.9068 kW/hour. Furthermore, the power plant operation can minimize the power loss by 39.9249 percent, from 36.2280 kW to 21.7640 kW.

Read more…

In this paper, an automatic method for detection of olive trees in RGB images acquired by an unmanned aerial vehicle (UAV) is developed. Presented approach is based on the implementation of RetinaNet model and DeepForest Phyton package. Due to fact that original (pretrained) model used in DeepForest package has been built on images of various types of trees but without images of olive trees, original model detection was unsatisfactory. Therefore, a new image dataset of olive trees was created using sets of images chosen from five olive groves. For neural network training, individual olive trees were manually labeled, and new models were generated. Each model has been trained on different set of images from selected olive groves. Pretrained model and new models were compared and evaluated for various test scenarios. Obtained results showed high precision and recall values of proposed approach and great improvement in performance compared to the pretrained model.

Read more…

Decision-makers must make a suitable sequence of decisions under uncertainty in a relatively long period for particular projects and situations. Conventional decision-making approaches under uncertainty are based on expected utility theory and do not sufficiently reflect the one-time nature of decisions. Similarly, the conventional approaches do not adequately incorporate the decision-maker’s intuitions in the decision-analysis process. Numerous studies have demonstrated that salience information (attention-grabbing) is crucial in human decision-making exercises. However, there is limited information on the decision-making approaches incorporating the salience information and the applications of such approaches in actual practice. This study applies an approach called the multistage one-shot decision-making approach (MOSDMA) to reevaluate a previous decision problem related to a department technology project from the sultanate of Oman. Unlike traditional lottery-based approaches, MOSDMA is scenario-based, introducing an essential alternative for multistage decision-making under uncertainty. The paper is the first contribution to using the passive focus point introduced in MOSDMA in actual applications. The aim is to verify the explicability and effectiveness of the suggested method for solving decision-making under uncertainty problems in actual practice. The paper exhibits positive findings and promising potential of the approach advocating further future studies in theory and application aspects.

Read more…

Traditional project planning in effort and duration estimation techniques remain low to medium accurate. This study seeks to develop a highly reliable and efficient hybrid Machine Learning model that can improve cost and duration prediction accuracy. This experiment compared the performance of five machine learning models across three different datasets and six performance indicators. Then the best model was verified with three other types of live project data. The results indicated that the MLR-DNN is a highly reliable, effective, consistent, and accurate machine learning model with a significant increase in accuracy over conventional predictive project management tools. The finding pointed out a potential gap in the relationship between dataset quality and the Machine Learning model’s performance.

Read more…

Determining the best candidates for a certain job rapidly has been one of the most interesting subjects for recruiters and companies due to high costs and times that takes the process. The accuracy of the models, particularly, is heavily influenced by the discriminant variables that are chosen for predicting the candidates scores. This study aims to develop an performance job prediction systems based on hybrid neural network and particle swarm optimisation which can improve recruitment screening by analyzing historical performances and conditions of em- ployees. The system is built in four stages: data collection, data preprocessing, model building and optimisation and finally model evaluation. Additionally, we highlight the significance of Particle Swarm Optimization (PSO) in enhancing the performance of the models created by presenting a training algorithm that uses PSO. We conduct a study to compare the performance of each hybrid model and summarize the results.

Read more…

This work investigated the coupling behavior of the scattered-bend loss in displacement sensor during the bending of the fiber by using a multimode polymer optical fiber (POF). To utilize the scattered-bend effect for displacement measurement, a side coupling technique can be used by twisting a pair of POF fibers and bent the structure into a loop. The working principle of the sensor is quite simple. The bent radius grows smaller as the fiber draughts which simulate a change of displacement. The scattered-bend loss increases as the illuminating fiber is bent in decreasing angle and the light being coupled to the receiving fiber. The fabricated sensor is tested based on static measurement analysis and the sensor is characterized by its sensitivity, resolution, linearity, and repeatability error. From the experiment, the fabricated sensor has a range of roughly 160 mm with a sensitivity of 0.817 nW/mm, a resolution of 1.228 mm, and a repeatability error of 1.856 %. The sensor exhibits high linearity from 0 mm to 80 mm. The sensor’s design structure and analysis are simple, comprehensive, and cost-effective, with potential benefits in industrial applications.

Read more…

Cyber security is based on different principles such as confidentiality and integrity of transmitted data. One of the main methods to send confidential messages is to use a shared secret to encrypt and decrypt them. Even if the amortized computational complexity of the hashing functions is Ο(1), there are several situations when it is not possible to use them due to the lack of computing power or the need to keep completely hidden the communication to other parties in the network. Covert Channels (CCs) are an excellent alternative in all these cases because they hide the private message in legitimate communication channels without the need to allocate additional resources to communicate. For this reason, they are difficult to identify because they are fully camouflaged in legitimate traffic. Unfortunately, CC technique is also used by hackers to exfiltrate network data and initiate cyber-attacks against devices in the system: Internet of Medical Things (IoMT) are one of the most vulnerable devices affected by this type of attack. It is therefore essential to create a system that can autonomously identify the presence of a malicious CCs to safeguard the health of patients. This paper describes an approach to create a Covert Timing Channel (CTC) based on TCP packets between client and server and how it is possible to detect the hidden communication using an innovative pipeline composed by several Machine Learning (ML) and Deep Learning (DL) models, such as Convolutional Neural Network (CNN), Siamese Neural Network (SNN) and K-Nearest Neighbors (K-NN). Considering 4 different message types exchanged in CTC, the proposed pipeline achieved 94% accuracy in identifying covert messages in the channel.

Read more…

Recently, the Multiple-input multiple-output (MIMO) antennas have been used a lot and attracted many researchers in advanced high-speed wireless communication systems. MIMO antennas are an essential part not only in access points but also in end-user devices. This technology allows a significant increase in channel capacity, but also lead to a challenge of minimize mutual coupling and in the meantime reserved antennas’ compact size. In this study, we propose a triple-band MIMO antenna design. By using a symmetric parasitic structure, isolation between radiation elements is significantly improved. Besides, each antenna element is designed using a combination of planar structure and 8 Fibonacci curves that makes it compact in size and easy to fabricate in the circuit board of 5G/ Wi-Fi 6 terminals. With a total dimension of 34.8* 68.2*1.6mm3, the proposed MIMO antenna design can operate at three bands of 2.4GHz, 3.5GHz, and 5GHz with wide bandwidths of 11.4%, 9.4%, and 14.58%, respectively. The results are analyzed based on simulation, measurement, and experiment.

Read more…