Volume 7, Issue 3

Overall Performance (OP) measurement is an essential instrument in sustainable manufacturing implementation and management. The effective use of key performance indicators (KPIs) can potentially contribute to identify the firm’s overall performance, provide the crucial gaps between desired results and current actions, and thus facilitate the implementation and execution of improvement strategies. This study attempts to give an insight into the perception of OP measurement and management in the automotive industry and explore the KPIs pertinent to this sector. This research is conducted in automobile organizations based in Morocco. We first carried out a literature review to determine the commonly used indicators of sustainability management in manufacturing. We then conducted a survey on a sample of firms to investigate the OP system management and how the proposed initial set of KPIs is perceived and used in the field. Findings reveal some management problems of the OP measurement system in the automobile sector. It was found that only a minority of companies use dedicated applications to manage their set of indicators. There is a lack of well-defined and standardized KPIs, which generally affect data quality. Moreover, most companies have a minimum percentage of decisions based on KPIs use, and only a few are satisfied with their overall performance measurement systems. However, analysis indicates no substantial differences in the perception of KPIs’ importance among various respondents. Results showed that the most used KPIs are perceived as the most important. Consequently, sixteen indicators under the three dimensions of sustainability were presented as KPIs for OP measurement in the automotive industry. These indicators will hopefully serve the development of an OP management tool to support sustainability in this sector. The study is evenly valuable for other developing countries as Morocco in sustainability implementation in the automobile field.

Read more…

Social networks have become a valuable platform for tracking and analyzing Internet users’ feelings. This analysis provides crucial information for decision-making in various areas, such as politics and marketing. In addition to this challenge and our interest in the field of big data and sentiment analysis in social networks, we have dedicated this work to combine different aspects of methods or techniques leading to the facilitation of feelings classification in social networks, including text analysis and sentiment analysis. We expose the approaches and the algorithms of supervised machine learning for the classification of feelings. We further our research to concisely present the methods of data representation and the parameters used to evaluate a sentiment analysis method in the context of social networks, with a section presenting our novel lexicon-based approach to give more accurate results in classifying Arabic text. The proposed approach has shown a promising accuracy percentage, especially the precision of the sentiment detected from text with F-Score up to 66%.

Read more…

The development of an aggregated system consisting of autonomously developed components is usually implemented as a self-contained unit. If such an aggregation is understood as a system of systems (SoS) that communicates via interfaces with its autonomous subsystems and components, the interfaces and communication exchange should play a central role in the architectural design. In fact, complete and exact interface specifications simplify loose coupling of independent systems into an aggregation. Since an SoS consists of variant and non-variant subsystems, the main challenge in SoS development is the identification of all true variants and its deviating attributes within an SoS. If the system variants are identified at an early stage of the development process, redundant work in the interface design can be substantially reduced. This paper presents an efficient method to identify SoS variants with regard to life cycle management and it shows how to configure a variant-oriented SoS with a standardized communication interface. For the development, the forward-looking model-based systems engineering approach is recommended to create executable specification parts and to detect errors early on through simulations.

Read more…

The main purpose in this work is to explore the fact that chaos, as a biological characteristic in the brain, should be used in an Artificial Neural Network (ANN) system. In fact, as long as chaos is present in brain functionalities, its properties need empirical investigations to show their potential to enhance accuracies in artificial neural network models. In this paper, we present brain-inspired neural network models applied as pattern recognition techniques first as an intelligent data processing module for an optoelectronic multi-wavelength biosensor, and second for breast cancer identification. To this purpose, the simultaneous use of three different neural network behaviors in the present work allows a performance differentiation between the pioneer classifier such as the multilayer perceptron employing the Resilient back Propagation (RProp) algorithm as a learning rule, a heteroassociative Bidirectional Associative Memory (BAM), and a Chaotic-BAM (CBAM). It is to be noted that this would be in two different multidimensional space problems. The later model is experimented on a set of different chaotic output maps before converging to the ANN model that remarkably leads to a perfect recognition for both real-life domains. Empirical exploration of chaotic properties on the memory-based models and their performances shows the ability of a specific modelisation of the whole system that totally satisfies the exigencies of a perfect pattern recognition performance. Accordingly, the experimental results revealed that, beyond chaos’ biological plausibility, the perfect accuracy obtained stems from the potential of chaos in the model: (1) the model offers the ability to learn categories by developing prototype representations from exposition to a limited set of exemplars because of its interesting capacity of generalization, and (2) it can generate perfect outputs from incomplete and noisy data since chaos makes the ANN system capable of being resilient to noise.

Read more…

Gamified Instructional Media has recently been widely used in the education sector to improve students’ abilities. Using Gamified Instructional Media at the elementary school level becomes more interesting because it is in accordance with the way children learn K1-K6. The research aims to identify the gamified instructional using Genially to improve students’ critical and creative thinking skills. A quasi-experimental method was applied using a nonequivalent control group research design. The research subject is 40 students of Public Primary School in Pekanbaru. The results show a significant effect of the gamified instructional learning using Genially toward students’ critical and creative thinking skills. Besides, there is a significant difference in students’ critical and creative thinking skills between the control and experimental group. This study implies that gamified instructional media with Genially can support teachers and teaching practices.

Read more…

Predicting the credit card default is important to the bank and other lenders. The credit default risk directly affects the interest charged to the borrower and the business decision of the lenders. However, very little research about this problem used the Generalized Linear Model (GLM). In this paper, we apply the GLM to predict the risk of the credit card default payment and compare it with a decision tree, a random forest algorithm. The AUC, advantages, and disadvantages of each of the three algorithms are discussed. We explain why the GLM is a better algorithm than the other two algorithms owing to its high accuracy, easy interpretability, and implementation.

Read more…

The use of decentralized reinforcement learning (RL) in the context of multi-agent systems (MAS) poses some difficult problems. The speed of the learning process for example. Indeed, if the convergence of these algorithms has been widely studied and mathematically proven, they suffer from being very slow. In this context, we propose to use RL in MAS in an intelligent way to speed up the learning process in these systems. The idea is to consider the MAS as a new environment to be explored and the communication, between the agents, is limited to the exchange of knowledge about the environment. The last agent to explore the environment has to communicate the new knowledge to the other agents, and the latter have to build their knowledge bases taking into account this knowledge. To validate our method, we chose to evaluate it in a grid environment. Agents must exchange their tables (Qtables) to facilitate better exploration. The simulation results show that the proposed method accelerates the learning process. Moreover, it allows each agent to reach its goal independently.

Read more…

DNN implementation and deployment is quite a challenge within a resource constrained environment on real-time embedded platforms. To attain the goal of DNN tailor made architecture deployment on a real-time embedded platform with limited hardware resources (low computational and memory resources) in comparison to a CPU or GPU based system, High Performance SqueezeNext (HPS) architecture was proposed. We propose and tailor made this architecture to be successfully deployed on Bluexbox 2.0 by NXP and also to be a DNN based on pytorch framework. High Performance SqueezeNext was inspired by SqueezeNet and SqueezeNext along with motivation derived from MobileNet architectures. High Performance SqueezeNext (HPS) achieved a model accuracy of 92.5% with 2.62MB model size at 16 seconds per epoch model using a NVIDIA based GPU system for training. It was trained and tested on various datasets such as CIFAR-10 and CIFAR-100 with no transfer learning. Thereafter, successfully deploying the proposed architecture on Bluebox 2.0, a real-time system developed by NXP with the assistance of RTMaps Remote Studio. The model accuracy results achieved were better than the existing CNN/DNN architectures model accuracies such as alexnet_tf (82% model accuracy), Maxout networks (90.65%), DCNN (89%), modified SqueezeNext (92.25%), Squeezed CNN (79.30%), MobileNet (76.7%) and an enhanced hybrid MobileNet (89.9%) with better model size. It was developed, modified and improved with the help of different optimizer implementations, hyper parameter tuning, tweaking, using no transfer learning approach and using in-place activation functions while maintaining decent accuracy.

Read more…

In Al_0.35Ga_0.65As/GaAs/Al_0.25Ga_0.75As quantum wells, the hole-confined polar optical phonon interaction is investigated. To calculate the valence band structure, we use the Luttinger-Kohn Hamiltonian with the k.p method. Within the dielectric continuum model, the hole-confined phonon scattering rates of intrasubband heavy holes in quantum well are calculated. It is found that the scattering rates are governed by an overlap integral and the density of states. Moreover, the scattering rates are reduced under compressive hydrostatic strain for low hole energy. The anisotropic effect on hole-confined phonon interaction is also studied.

Read more…

A novel PVTL (Process, Voltage, Temperature, Leakage) detection circuit consisting of four individual detectors is proposed in the investigation. Voltage Variation Detector is composed of a feedback control block comprising multi-stage delay cells using high Vth devices such that 0.5% of VDD variation can be detected. Temperature Detector based on a current to pulse converter is proved to attain high linearity of temperature sensing. PMOS Variation Detector and NMOS Variation Detector are carried out using threshold voltage sensors and ring oscillators, respectively. Thus, all process corners can be clearly differentiated using pulse counts. Leakage Detector is realized by a single-MOSFET leakage current detector. Most of prior leakage detectors compensate for leakage current instead of detecting the precise amount of the leakage current. The proposed leakage detector, however, can accurately detect the leakage current of CMOS transistors, where a Strobe pulse generator is used as a detection switch. Thus, the detection time is predictable. It elevates the reliability of the detection result. The proposed PVTL detector design is implemented using a typical 180 nm CMOS process to justify the performance. Measurement shows that the proposed design is the best of all prior PVTL detectors in terms of accuracy.

Read more…

The Internet of Things (IoT) is the evolving paradigm of interconnectedness of objects with varied architectures and resources to provide ubiquitous and desired services. The popularization of IoT-connected devices facilitating evolution of IoT applications does come with security challenges. The IoT with the integration of wireless sensor networks possess a number of unique characteristics, so the implementation of security in such a restrictive environment is a challenging task. Due to the perception that security is expensive in terms of computation, power and user-interface components, and as sensor nodes or low-power IoT objects have limited resources, it is desired to design security mechanisms especially routing protocols that are light weighted. Bio-inspired mechanisms are shown to be adaptive to environmental variations, robust and scalable, and require less computational and energy resources for designing secure routing algorithms for distributed optimization. In IoT network, the malicious intruders can exploit the routing system of the standardized routing protocol, e.g., RPL (The Routing Protocol for Low-Power and Lossy Networks), that does not observe the node’s routing behavior prior to data forwarding, and can launch various forms of routing attacks. To secure IoT networks from routing attacks, a secure trust aware ACO-based WSN routing protocol for IoT is proposed here that establishes secure routing with trustworthy nodes. The trust evaluation system, is enhanced to evaluate the node trust value, identify sensor node misbehavior, and maximize energy conservation. The performance of the proposed routing algorithm is demonstrated through MATLAB. Based on the proposed system, to find the secure and optimal path while aiming at providing trust in IoT environment, the average energy consumption is minimized by nearly 50% even as the number of nodes has increased, as compared with the conventional ACO algorithm, a current ant-based routing algorithm for IoT-communication, and a present routing protocol RPL for IoT.

Read more…

Technologies designed for digital provenance, especially the Internet of Things (IoT) and blockchain, may allow for security, transparency, and traceability in the global supply chain. However, upstream nodes in the supply chain that work for large-scale production suppliers are not considered. In addition, most IoT blockchain systems adopt an ID-based signature scheme that may affect the efficiency of IoT devices. We propose using aggregate verification to improve the security and efficiency of ID-based verification, reduce network traffic on the blockchain, and transfer computing overhead to aggregator nodes. This paper implements a multi-layer blockchain for Agriculture 4.0 supply chain management that has higher efficiency, effectiveness, and security in comparison to conventional blockchains. We design a Multi-Layer Aggregate Verification (MLAV) solution to improve supply chain management with IoT Blockchain for Agriculture 4.0 through the following methods. First, we use a multi-layer IoT blockchain system to reduce Ethereum gas fee. Second, we design an ID-based Aggregate Verification scheme, thereby eliminating the certificate management cost in the traditional Public Key Infrastructure (PKI) and reducing bandwidth and computation time requirements. Third, we implement a three-layer blockchain infrastructure. In Layer 1, IoT devices sense and upload data to the system’s database; in Layer 2, smart contracts execute aggregate ID-based signature verification from IoT devices and upload the transactions to the private blockchain; in Layer 3, a batch converts the layer 2 data and uploads its Merkle root to Ethereum, thereby reducing the required gas fee.

Read more…

Nigeria has one of the highest deforestation rates in the world, due mainly to felling of trees for fuelwood and charcoal production. This challenge could be managed if agricultural waste briquettes are used to augment the fuelwood demand for cooking energy provisioning. Energy density of biomass fuels can be raised by blending with coal, but particulate matter emissions are also increased in the process. Therefore, some physical and fuel properties of briquettes produced from blends of corn cob and palm mesocarp fibre (PMF) with Sub-bituminous coal from Onyeama mine, using cassava starch binder were studied. Blending ratios were varied from 0% to 100% biomass. After briquetting in a hydraulic press, drying and characterization, Water Boiling Test (WBT) of the briquette samples was performed using Laboratory Emissions Measuring System (LEMS). Results showed that average High-Power Thermal Efficiency was 28.8% for corn cob/coal and 26.0% for PMF/coal briquettes, but High-Power CO emissions decreased from 10.8 mg/MJd to 8.9 mg/MJd (8.31-6.90 ppm) as the composition of corn cob increased from 0% to 100%. Corn cob/coal briquettes produced lower PM emissions than pmf/coal, although both were above the WHO recommended limits.

Read more…

The NFC wireless charging feature is an extension of the NFC technology that can be implanted on wearables. The purpose of this paper is to show how to increase power transfer efficiency on both transmitter and receiver antenna systems. To demonstrate this problematic, firstly this paper gives an overview of how this NFC feature is implemented (architecture, power transfer, carrier frequency, communication bandwidth…), and can be complementary to Qi technology. Then, it provides a study on how to improve the power transfer efficiency on the antennas. To perform this result, the designer can adapt some antenna parameters as coupling coefficient, quality factor, matching method, and the antenna size. If these recommendations are respected, the power transfer efficiency between the antennas could reach between 70% and 80% with the NFC charging technology.

Read more…

In this paper, we propose a highly accurate indoor position and direction estimation system using a simple fully connected deep neural network (DNN) model on Bluetooth Low Energy (BLE) Received Signal Strength Indicators (RSSIs). Since the system’s ultimate goal is to function as an indoor navigation system, the system estimates the indoor position simultaneously as the direction of movement using BLE RSSI fingerprints recorded indoors. To identify the direction of movement along with the position, we decided to use multiple time instances of RSSI measurements and fed them to a fully-connected DNN. The DNN is configured to output the direction with the location simultaneously. RSSIs are known to be affected by various fluctuating factors in the environment and thus tend to vary widely. To achieve stable positioning, we examine and compare the effects of temporal interpolation and extrapolation as preprocessing of multiple RSSI sequences on the accuracy of the estimated coordinates and direction. We will also examine the number of beacons and their placement patterns required for satisfactory estimation accuracy. These experiments show that the RSSI preprocessing method optimum for practical use is interpolation and that the number and placement of beacons to be installed does affect the estimation accuracy significantly. We showed that there is a minimum number of beacons required to cover the room in which to detect the location if the estimation error is to be minimized, in terms of both location and direction of movement. We were able to achieve location estimation with an estimation error of about 0.33 m, and a movement estimation error of about 10 degrees in our experimental setup, which proves the feasibility of our proposed system. We believe this level of accuracy is one of the highest, even among methods that use RSSI fingerprints.

Read more…

Dust is ubiquitous in our daily environment—outdoor and indoor. In modern times, people often spend the majority of their time at home, in offices, at work or in schools. Suspended particles such as tiny crumbs up to long fibers generate indoor dust deposits. Inhouse sources are the interior releasing abraded fibers from carpets, bedding and clothing as well as the human itself distributing skin cells, lost hairs and food residues. External sources are finest sand, pollen, exhaust particles and microorganisms (e.g., dust mites). An exposure to heavy metals in certain concentrations may affect the human health and may lead to intoxication, allergies or carcinogenic effects. The heavy metals amount in indoor dust depends on the environmental conditions, requiring a sampling with adequate sampling points and numbers. High sampling numbers ensure good coverage of the area to be examined. Therefore, fast and reliable measurement methods for identification and quantification of the elemental composition are needed. To meet these requirements, a robotic system for automated sample preparation and determination of heavy metals in indoor dust using ICP-MS was developed. The values for repeatability, recovery rate, within-laboratory precision, measurement precision and the limits of detection and quantification were determined for both, the manual and the automated process. Furthermore, the individual process steps and times were compared. Besides the processing of certified reference material, inhouse dust samples from different origin were prepared and measured to give a first overview of the inhouse dust composition. The results show, that the robot-assisted system is well-suited for the heavy metal screening in indoor dust.

Read more…

Information security behaviour is an integral part of modern business and has become a central theme in many research studies. One of the essential tools available that can be used to influence information security behaviour is information security policies (ISPs). These types of policies, which is mandatory in most organisations, are formalised rules and regulations which guide the safeguarding of information assets. Despite a significant number of ISP and related studies, a growing number of studies report ISP non-compliance as one of the main factors contributing to undesirable information security behaviour. It is noteworthy that these studies generally do not focus on the opinion of users or employees about the contents of the ISPs that they have to adhere to. The traditional approach to obtain user or employee opinions is to conduct a survey and ask for their opinion. However, surveys present unique challenges in fake answers and response bias, often rendering results unreliable and useless. This paper proposes a deep learning affective computing approach to perform sentiment analysis based on facial expressions. The aim is to address the problem of response bias that may occur during an opinion survey and provide decision-makers with a tool and methodology to evaluate the quality of their ISPs. The proposed affective computing methodology produced positive results in an experimental case study. The deep learning model accurately classified positive, negative, and neutral opinions based on the sentiment conveyed through facial expressions.

Read more…

Optoelectronic devices applications based on Organic–inorganic perovskites are promising and effective low-cost energy materials due to their exceptional physical properties which include high carrier mobility, high optical absorption coefficient, and long carrier diffusion length. In the presented work, a TiO2/NiO+5% Fe quantum dots (QDs)–doped CsMAFAPbI2Br2 perovskite heterojunction broadband photodetector was fabricated on FTO/glass substrate. The photodetector can detect a wide range of wavelengths, from UV to Vis (100–800 nm), has high responsivity (0.99 A/W), and has excellent detectivity (8.9 × 1012 Jones). Scanning electron microscopy (SEM), transmission electron microscopy (TEM), X-ray diffraction (XRD), atomic force microscopy (AFM), and electrical (I–V) characterization were used to measure the responsivity and detectivity of the photodetectors. Doping with NiO+5% Fe QDs protected the device from oxygen and moisture and improved the morphology of the perovskite by reducing pit defects. The results showed high performance and the potential of a NiO+5% Fe QDs–doped triple cation perovskite photodetector device.

Read more…

Building detection in aerial or satellite imagery is one of the most challenging tasks due to the variety of shapes, sizes, colors, and textures of man-made objects. To this end, in this paper, we propose a novel approach to extracting buildings in high-resolution images based on prior knowledge of the shadow position. Firstly, the image is split into superpixel patches; the colors and texture features are extracted for those patches. Then using the machine learning method (SVM), four classes are made: buildings, roads, trees, and shadows. According to the prior knowledge of shadows position, a seed point initial has been defined along with an adaptive regional growth method to determine the approximate building location. Finally, applying a contouring process included an open morphological operation to extract the final shape of buildings. The performance is tested on aerial images from New Zealand area. The proposed approach demonstrated higher detection rate precision than other related works, exceeding 97% despite the complexity of scenes.

Read more…

Redundant robotic systems provide great challenges in solving kinematics and control problems, that are yet also open opportunities for exploring new, diverse and intelligent ideas and methods. In this paper, an advanced control method is proposed for position control problems of redundant robots with output constraints. The controller is structured with two control layers. In the high-level control layer, a cost function is first synthesized from the main control objective under constraint conditions. Virtual control signals are then reckoned to optimize the cost function using a soft Momentum-Levenberg-Marquardt approach. To realize the high-level control command, a nonlinear control signal is employed in the low-level control layer throughout a new nonsingular terminal sliding mode control structure. Comparative simulation results verified on a 7-DOF robotic arm model confirmed the effectiveness of the proposed control algorithm.

Read more…