Search

Achieving robust cross contexts speech emotion recognition (SER) has become a critical next direction of research for wide adoption of SER technology. The core challenge is in the large variability of affective speech that is highly contextualized. Prior works have worked on this as a transfer learning problem that mostly focuses on developing domain adaptation strategy. However, many of the existing speech emotion corpora, even those considered as large scale, are still limited in size resulting in an unsatisfactory transfer result. On the other hand, directly collecting context-specific corpus often results in an even smaller data size leading to an inevitably non-robust accuracy. In order to mitigate this issue, we propose the concept of enhancing the affect-related variability when learning the in-context acoustic latent representation by integrating out-of-context emotion data. Specifically, we utilize adversarial autoencoder network as our backbone with multiple out-of-context emotion labels derived for each in-context samples that serve as an auxiliary constraint in learning the latent representation. We extensively evaluate our framework using three in-context databases with three out-of-context databases. In this work, we demonstrate not only an improved recognition accuracy but also a comprehensive analysis on the effectiveness of this representation learning strategy.

With the massive deployment of the Internet of things (IoT) devices, many data analysis applications emerge for the large amount of data accumulated by IoT. Federated learning (FedL) on IoT devices is an appealing mode to train a precise data analysis model. However, existing federated learning schemes either take expensive computation costs (e.g., public key cryptographic operations) or a large number of interactions among participants. Obviously, these schemes are unsuitable for IoT devices due to the limited computational and communication resources. In this work, we propose a lightweight privacy-preserving federated learning scheme for IoT devices. To protect the privacy of individual local data, we add masks to intervening parameters. An effective secret sharing scheme is adopted to ensure that masks can be eliminated accurately. Considering that FedL involves multiple iterations and mask generation for each iteration costs a large number of interactions among users for privacy guarantee, we also design a secure mask reusing mechanism for large-scale FedL tasks. We prove that our scheme is secure against the honest-but-curious model. In addition, we also expand our scheme to deal with the collusion attack. Extensive experiments on real IoT devices demonstrate the accuracy and efficiency of our work.

In Ancient days there is no machine is used for digging the soil for plantation. The digging of soil for plantation is done manually by humans with the help of the hand auger and it takes more time and requires humans in large scale. This type of traditional process is very hard and it requires a high amount of human efforts. Due to these reasons, nowadays soil digging machines are used in agricultural field for decreasing the human efforts and do the work in efficient manner. The utilization of soil drilling machines in the agricultural fields takes a wide place. These machines are available in the market with different capacities. If the capacity of machine increases automatically cost of machine is also increases. At present in market soil drilling machines is available with diesel engines which are high in cost and heavier in weight. The aim of our project is to reduce the cost of the machine and make it as portable for easy carrying. So due to these reasons we are using petrol engine instead of diesel engine and we attach a iron frame to it for making it portable. This project involves design and fabrication of soil drilling machine. The design of soil drilling machine is done by NX 12 and CATIA V 5 R19 and the fabrication is completed by means of various fabrication techniques.

With the development of mobile networks and intelligent equipment, as a new intelligent data sensing paradigm in large-scale sensor applications such as the industrial Internet of Things (IIoT), mobile crowd sensing (MCS) assigns industrial sensing tasks to workers for data collection and sharing, which has created a bright future for building a strong industrial system and improving industrial services. How to design an effective worker selection mechanism to maximize the utility of crowdsourcing is the research hotspot of mobile sensing technologies. This paper studies the problem of least workers selection to make large MCS system perform sensing tasks more effective and achieve certain coverage with certain constraints being meeting. A many-objective worker selection method (MaOWS) is proposed to achieve the desired tradeoff and an optimization mechanism is designed based on the enhanced differential evolution algorithm (EDEA) to ensure data integrity and search solution optimality. The effectiveness of the proposed method is verified through a large scale of experimental evaluation data sets collected from real world.

Today our society needs more energy for day-to-day activities due to rapid globalization and industrialization. In order to minimize the stress and dependency on fossil fuel, the most sustainable way is to harness suns energy. Solar energy is characterized by low cost, environment friendly, does not require frequent maintenance and most importantly, negligible maintenance cost. However, there is a requirement of monitoring solar installation in order to maximize the output power by setting real-time angles with the suns position. This can be easily done by the adoption of IoT technologies. The Internet of Things is often utilized in the measurement of solar energy for efficiency. It's also utilized to maintain the solar plant's health. The cost of renewable energy technology is decreasing throughout the globe, which encourages the construction of large-scale solar plants. Automation of plant observation on such a large scale of preparatory deployments requires sophisticated systems reliant on Internet connections since the majority of field units are situated in isolated, remote locations and therefore are not overseen from a central office. The project is based on the use of the most up-to-date, cost-effective method for remotely monitoring a solar plant performance by the inclusion of IoT. It can assist with plant maintenance, problem diagnostics, and real-time monitoring.

Motivation: The prediction of drug-target interaction (DTI) from heterogeneous biological data is critical to predict drugs and therapeutic targets for known diseases such as tumor and bowel disease. The study of DTI based on drug representation learning can strengthen or integrate our knowledge of pharmacological and chemical phenomena. Therefore, there is a strong motivation to develop effective methods that can detect these potential drug-target interactions. Results: We have developed a novel Knowledge-Enhanced Multi-View framework (KEMV) to predict unknown DTIs from pharmacological data and chemical data on a large scale. The proposed method consists of two steps: (i) learning more comprehensive drug representations via the proposed multi-view attention mechanism, which bridges pharmacological and chemical information, and interactively summarizes the attention values depending on varying interactions between different pairs of drug features. (ii) predicting unknown drug-target interactions based on the drug and target representations. The method is tested on real-world dataset KEGG with three classes of important drug–target interactions involving enzymes, ion channels, and G-protein-coupled receptors. Our framework is proven to uncover potential DTIs with scientific evidences explaining the mechanism of the interactions through the processing of high-dimensional, heterogeneous, and sparse drug data. Availability: Our code is available at: https://github.com/YuanKQ/DTI-Prediction

In hazardous gas leakage accidents in chemical clusters, quickly and accurately locating the diffusion source via mobile sensors (searcher) can effectively speed up emergency response. While state-of-the-art researches designed kinds of autonomous sensing and searching techniques, e.g. cognitive-based strategies to complete the localization of the leakage source; the current studies cannot meet chemical cluster’s requirements with complex road network constraints and spatial scales. To address this problem, we firstly combine a cognitive search strategy with great theoretical performance (Entrotaxis algorithm) with an intermittent search strategy and propose a fresh searching algorithm, namely Entrotaxis-Turn (ET). The hybrid algorithm can make use of the triggered turn motion to bypass obstacles to avoid falling into a loop, thereby improving the success rate and saving the searching time. Secondly, we extend the ET algorithm to the multi-robot collaboration pattern (Multi-ET algorithm), in which each mobile sensor shares the information collected along its route with each other. To reveal the performance of the proposed algorithms, a simulation scene is built in a typical chemical cluster scene and diffused gas is generated based on the advection-diffusion equation. Through Monte Carlo simulations, the optimal parameter combinations in the ET and Multi-ET algorithm are obtained. Finally, a verification experiment is carried out based on the simulation scene generated by the computational fluid dynamics (CFD) model. Results show that compared with the Entrotaxis algorithm, the efficiency and success rate of the ET algorithm are dramatically improved. Moreover, the Multi-ET algorithm obtains the optimal source searching performance.

Molecular dynamics simulations are an invaluable tool in numerous scientific fields. However, the ubiquitous classical force fields cannot describe reactive systems, and quantum molecular dynamics are too computationally demanding to treat large systems or long timescales. Reactive force fields based on physics or machine learning can be used to bridge the gap in time and length scales, but these force fields require substantial effort to construct and are highly specific to given chemical composition and application. The extreme flexibility of machine learning models promises to yield reactive force fields that provide a more general description of chemical bonding. However, a significant limitation of machine learning models is the use of element-specific features, leading to models that scale poorly with the number of elements. This work introduces the Gaussian multi-pole (GMP) featurization scheme that utilizes physically-relevant multi-pole expansions of the electron density around atoms to yield feature vectors that interpolate between element types and have a fixed dimension regardless of the number of elements present. We combine GMP with neural networks to directly compare it to the widely-used Behler-Parinello symmetry functions for the MD17 dataset, revealing that it exhibits improved accuracy and computational efficiency. Further, we demonstrate that GMP-based models can achieve chemical accuracy for the QM9 dataset, and their accuracy remains reasonable even when extrapolating to new elements. Finally, we test GMP-based models for the Open Catalysis Project (OCP) dataset, revealing comparable performance and improved learning rates when compared to graph convolutional deep learning models. The results indicate that this featurization scheme fills a critical gap in the construction of efficient and transferable reactive force fields.

Active electricity distribution systems (AEDS) play an important role on the social welfare that is highly relies on the reliability of these systems. In this regard, maintenance management tries to decrease component failures via repairing component that are more possible to be failed in operation time. In this manuscript, a stochastic MINLP maintenance scheduling model is proposed in which the reliability of AEDS's equipment is considered through employing a three-state Markov model. This model helps us to embed the uncertainty of AEDS's equipment service restoration time by considering online monitoring system for it and FLISR functions in AEDS. In this paper, minimizing the total cost of AEDS is determined as objective function considering total customer interruption cost, total cost of energy not supplied, total cost of prevented maintenance actions, total corrective maintenance actions cost implemented in the AEDS and total cost of implementing online monitoring for AEDS's component in maintenance scheduling. The proposed framework is applied in two different cases, Case I is maintenance scheduling without online monitoring system and Case II is maintenance scheduling with online monitoring system. The aforementioned cases are implemented in an IEEE Test System as well as a real scale distribution network. Moreover, several sensitivity analyses are implemented for indicating the efficiency of the proposed model.

Hydrogen is widely used in many industries, yet its role in the clean energy transition goes beyond being an element of these industries. Near-term practical large-scale clean hydrogen production can be made available by involving nuclear, solar, and other renewable energy sources in the process of hydrogen production, and coupling their energy systems to sustainable carbon-free hydrogen technologies. This requires further investigation and assessment of the different alternatives to achieve clean hydrogen using these pathways. This paper assesses the technoeconomics of promising hydrogen technologies that can be coupled to nuclear and solar energy systems for large-scale hydrogen production. It also provides an overview of the design, status and advances of these technologies.

Memetic algorithm is widely applied to optimise routing problems as it provides one way to combine local search with global search. However, the local search in memetic algorithm needs to be carefully designed according to the problem’s characteristics. In this paper, we consider a real-world large-scale waste collection problem with multiple depots, multiple disposal facilities, multiple trips and working time constraints. Vehicles with a limited capacity and working time can start from different depots, collect waste at different sites, make multiple trips to different disposal facilities to empty the waste and return to its origin. While existing work considered problems with multiple trips and time constraints, none have tackled problems with multiple depots, multiple disposal facilities, multiple trips as well as working time constraints. The change from “single-depot” to “multi-depot” not only reflects better the situation in real life but also leads to a qualitative different and more complex problem. In this paper, we first model this complex problem mathematically. Then, a novel region-focused memetic algorithm is proposed to tackle this new challenge. Compared to classic memetic algorithm, this region-focused one is enhanced by two major components: (i) a new heuristic-assisted solution initialisation algorithm and (ii) a region-focused local search with novel heuristics. Comprehensive computational studies show that our proposed approaches significantly outperform several state-of-the-arts on our real problem of thousands of tasks. New local search procedure and solution initialisation method significantly improve the search ability in combination with global search ability of memetic algorithm.

Hydrated cement is one of the complex composite systems due to the presence of multi-scale phases with varying morphologies. Calcium silicate hydrate (C–S–H), which is the principal binder phase in the hydrated cement, is responsible for the stiffness, strength, and durability of Portland cement concrete. To understand the mechanical and durability behavior of concrete, it is important to investigate the interactions of hydrated cement phases with other materials at the nanoscale. In this regard, the molecular simulation of cement-based materials is an effective approach to study the properties and interactions of the cement system at the fundamental scale. Recently, many studies have been published regarding atomistic simulations to investigate the cement phases to define/explain the microscopic physical and chemical properties, thereby improving the macroscopic performance of hardened binders. The research in molecular simulation of cementitious systems involves researchers with multidisciplinary backgrounds, mainly in two areas: (1) cement chemistry, where the hydration reactions govern most of the chemical and physical properties at the atomic scale; and (2) computational materials science and engineering, where the bottom-up approach is required. The latter approach is still in its infancy, and as such, a study of the prevailing knowledge is useful, namely through an exhaustive literature review. This state-of-the-art report provides a comprehensive survey on studies that were conducted in this area and cites the important findings.

In an environment with multiple heterogeneous multiservers across multiple clouds and data centers, optimal workload management and server speed setting makes strong impact on the aggregated performance, cost, and profit of a cloud service provider. Such a situation provides us a more challenging opportunity to address and discuss profit maximization of a service provider in a wider scale and to generate more significant influence. In this paper, we investigate profit maximization by optimal workload management and server speed setting for multiple heterogeneous multiservers in a federated cloud or a geo-distributed data center. The heterogeneous multiservers have different sizes, speeds, power consumption models, workload, performance, costs, and profit. To conduct rigorous study, each multiserver system is modeled by an M/M/m queueing system, such that the profit of a multiserver system can be characterized analytically. We address two problems, i.e., workload management without server speed setting and workload management with server speed setting. Both problems are formulated as multi-variable optimization problems. We develop numerical algorithms to solve these problems. We also provide numerical data for the purpose of illustration. This is the first paper which analytically discusses profit maximization for multiple heterogeneous multiservers in a federated cloud or a geo-distributed data center.

Mathematical optimization problems including a time dimension abound. For example, logistics, process optimization and production planning tasks must often be optimized for a range of time periods. Usually, these problems incorporating time structure are very large and cannot be solved to global optimality by modern solvers within a reasonable period of time. Therefore, the so-called rolling-horizon approach is often adopted. This approach aims to solve the problem periodically, including additional information from proximately following periods. In this paper, we first investigate several drawbacks of this approach and develop an algorithm that compensates for these drawbacks both theoretically and practically. As a result, the rolling-horizon decomposition methodology is adjusted to enable large scale optimization problems to be solved efficiently. In addition, we introduce conditions that guarantee the quality of the solutions. We further demonstrate the applicability of the method to a variety of challenging optimization problems. We substantiate the findings with computational studies on the lot-sizing problem in production planning, as well as for large-scale real-world instances of the tail-assignment problem in aircraft management. It proves possible to solve large-scale realistic tail-assignment instances efficiently, leading to solutions that are at most a few percent away from a globally optimal solution.

Exponentially growing number of workloads has been transferred from on-premises to the cloud environment during last decade. It incurs a constantly increasing load on the monitoring and troubleshooting capabilities of the platforms that host those applications nowadays. Data mining techniques and use of telemetry data have recently become an unavoidable means for tracking the behavior of cloud services. The research effort elaborated in this paper is focused on exploiting real-world data to build an automatic database troubleshooting system that exploits the combination of the more comprehensive statistical data science models and an expert system to perform the root cause analysis. An extensive evaluation study was conducted during the eight-month period with a plethora of Azure SQL production workloads for a varying number of databases. The obtained results confirmed the viability and cost-effectiveness of such an approach at the scale of the cloud.

Due to the rapid paradigm shift in IoT networks from wired and centralized to flexible wireless and decentralized networks, building effective and reliable distributed consensus mechanisms over wireless is becoming essential. Especially, since the performance of consensus over communication endpoints in a large-scale wireless network is limited by their communication capability, it requires a careful co-design of communication and consensus to attain a fast and scalable distributed wireless consensus mechanism with high resiliency against faulty nodes. Within this context, this paper addresses such problem by designing two wireless consensus mechanisms that well-suit in large-scale wireless networks. On the one hand, as a reinterpretation of the conventional referendum consensus (RC) in a large-scale wireless network, gossip-broadcasting based referendum consensus (GB-RC) is proposed. On the other hand, to overcome the scalability issue of the GB-RC, cooperative-broadcast based electoral-college consensus (CB-EC) is proposed. By mathematically analyzing the performance of both of the consensus mechanisms, in terms of consensus latency and resiliency against the faulty nodes, we show that the GB-RC outperforms the conventional RC, while the CB-EC significantly reduces the consensus latency compromising the stochastic resiliency. We further evaluate their performance numerically to show their effectiveness and feasibility under realistic large-scale wireless environments.

As the number of people involved in a decision-making problem increases, the complexity of the group decision-making (GDM) process increases accordingly. The size of participants and the heterogeneous information have important effects on the consensus reaching process in GDM. To deal with these two issues, traditional methods divide large groups into smaller ones to reduce the scale of GDM and translate heterogeneous information into a uniform format to handle the heterogeneity problem. These methods face two challenges: 1) how to determine the appropriate group size? and 2) how to avoid or reduce loss of information during the transformation process? To address these two challenges, this article uses fuzzy cluster analysis to integrate heterogeneous information for large-scale GDM problems. First, a large group is divided into smaller ones using fuzzy cluster analysis and the F-statistic is applied to determine the satisfactory number of clusters. The original information is retained based on the similarity degree. Then, a consensus reaching process is conducted within these small groups to form a unified opinion. A feedback mechanism is developed to adjust the small GDM matrix when any group cannot reach a consensus, and the heterogeneous technique for order preference by similarity to an ideal solution (TOPSIS) is used to select the best alternative. To validate the proposed approach, an experiment study is conducted using a practical example of selecting the best rescue plan in an emergency situation. The result shows that the proposed approach helps to choose the best rescue plan faster.

The landfill leachate is considered a toxic effluent composed of recalcitrant contaminants that requires innovative alternatives for its decontamination. Coupling between advanced oxidation processes (AOPs) and aerobic biological treatments are highlighted in this research. Therefore, a bibliographic review of the research made from 2010 to 2021 was developed. These combined alternatives were applied in leachates, and it is oriented towards the analysis of knowledge gaps, trends, and future proposals of the treatment combined that contribute to researchers who wish to work on the subject. These kinds of treatments were chosen due to a bibliometric analysis made. Also, the information was searched in several scientific database. This work was found to be unpublished, as no reviews were found so far that agglomerate studies of coupling between photocatalytic and aerobic biological processes to treat leachates. Besides, AOPs are ideal for treating wastewater of complex composition, however, when it is used as the only treatment, they are usually unprofitable, which justifies their coupling with biological treatments. Subsequently, it was determined that the knowledge main gap is the lack of documentation of treatment costs, which makes it difficult to implement on a real scale. In addition to this, the couplings trends are towards doping with metallic and non-metallic ions of the catalyst used in the photocatalytic process to improve the efficiency of these. Finally, future research should work on finding alternatives that allow the optimization of the resources used in the combined systems and on promoting the recovery of existing products in the leachate.

Classroom observation is an essential component to improve the quality of teaching and develop educational research. Nonetheless, traditional observation procedures involve previously trained observers, rendering them expensive and time-consuming. Thus, there is a need for a tool which enables to analyze multiple lessons in short time, represent those lessons consistently, and describe them according to the teaching strategies used over time to give teachers timely and continuous feedback. In this work, we propose two unsupervised approaches to represent the evolution of lessons based on topic modeling: textbook- and BERT-based. This way, we can improve the generalization capability of these tools and avoid the need of previously annotated data. Subsequently, we collapsed the derived topics into two super-topics to ease interpretability, and clustered the time series they define to characterize them according to the teaching strategies used. We applied our methodology to 195 science lessons revealing three patterns related to different strategies. Lastly, we compare the two approaches using different metrics for clustering assessment and discuss their benefits and detriments. We expect these models can become useful tools for teachers to obtain an automated and objective representation of their lessons, but also to other stakeholders who want to analyze teaching practices at a large scale.

In most of the developing countries, the economy is largely based on agriculture. The poor availability of skilled personnel and of appropriate supporting infrastructure, make crop fields vulnerable to the outbreak of plant diseases, possibly due to spreading viruses and fungi, or to adverse environmental conditions, such as drought. The mobile application PlantVillage Nuru, provides an invaluable tool for early detection of plant diseases and sustainable food production. A mobile device endowed with Nuru is a powerful mobile sensor: it analyzes plant images and uses an AI engine to recognize health issues. In this paper we propose a crowd-sensing framework, where Nuru is adopted at large scale in the farmer population. We tackle the device deployment problem, where device mobility is only partially controllable, mostly in an indirect manner, through incentives. We propose two problem formulations, and related algorithms, to minimize the number of required smartphones while providing sufficient geographical coverage. We study the proposed models in simulated as well as real scenarios, showing that they outperform current solutions in terms of monitoring accuracy and completeness, with lower cost. Then we describe the test-bed implementation, confirming the applicability of the proposed crowd-sensing framework in a real scenario in Kenya.

Human-oriented and machine-generated data in cyber-physical-social systems are often complicated graph-structured. Graph-powered learning methods are conducive to discovering valuable knowledge from large scale graph data and improving decision making processes. However, due to the neglect of diverse relations among things, most existing knowledge reasoning studies are inherently flawed and inefficient in processing the heterogeneous graphs with high-order connectivity. Tensor, as a powerful and effective tool to model high-level semantic interactions between various things, can provide high-order IoT graph with new perspectives and possibilities. Therefore, this paper innovatively proposes a collaborative Artificial Intelligence of Things data analysis and application framework based on Boolean tensors, which supports the expression and fusion of heterogeneous graph and ultimately promotes the AI processing. In this context, we focus on developing an incremental Boolean tensor factorization (IBTF) approach for efficient knowledge reasoning to meet the requirements of real-time and high level quality demands for intelligent services. To the best of our knowledge, we are the first to do this work. More concretely, we present factors update and binary features merge algorithms for the integrated graph tensors to avoid numerous repeated calculations of historical data. Experimental results on general synthetic data sets demonstrate that the IBTF approach proposed in this paper guarantees nearly equal approximate accuracy while reducing execution time by dozens and even more of times. Furthermore, experimental evaluations and interpretability analysis on real-world datasets verify the practicality of the proposed framework and approach.

Interconnection networks play an essential role in the architecture of high-performance computing (HPC) systems. In this article, we explore the Galaxyfly family to build flexible-scale interconnection networks. Galaxyfly is guaranteed to retain a small constant diameter while achieving a flexible tradeoff between network scale and bisection bandwidth. Galaxyfly not only supports small-scale interconnection networks with smaller diameter but also lowers the demands for high-radix routers and is able to utilize routers with moderate radix to build exascale interconnection networks. We analyze the constructible configuration of Galaxyfly and evaluate the properties of Galaxyfly. We conduct extensive simulations and analysis to evaluate the performance, cost, and power consumption of Galaxyfly on physical layout against state-of-the-art topologies. The results show that our design achieves better performance than most existing topologies under typical HPC workloads, and is cost-effective to deploy for exascale HPC systems.

International audience; Optimizing nuclear unit outages is of significant economic importance for the French electricity company EDF, as these outages induce a substitute production by other more expensive means to fulfill electricity demand. This problem is quite challenging given the specific operating constraints of nuclear units, the stochasticity of both the demand and non-nuclear units availability, and the scale of the instances. To tackle these difficulties we use a combined decomposition approach. The operating constraints of the nuclear units are built into a Dantzig-Wolfe pricing subproblem whose solutions define the columns of a demand covering formulation. The scenarios of demand and non-nuclear units availability are handled in a Benders decomposition. Our approach is shown to scale up to the real-life instances of the French nuclear fleet.

Currently, there is a beginning of liquefied natural gas (LNG) market transformation, which is deepening global supply chains through the growth of the number of small-scale projects. The main driver of this transformation was the creation of large-scale LNG hubs, which did not meet expectations for improving the efficiency of gas distribution systems (like in Europe). This has led to an understanding of the need of increasing flexibility and decentralization of gas supply projects, which is almost impossible to implement with traditional large and medium scale facilities. Increasing decentralization leads to a change in the role and composition of stakeholders, and, consequently, determines new algorithms and schemes for their interaction, which are poorly described in the scientific literature. This paper systematizes information about the specifics of such project implementation, defines a list of key stakeholders, their area of responsibility and area of interests. Proposed conceptual model of interaction between stakeholders has shown a significant dependence of small-scale projects on regional economic and infrastructure conditions, as well as on the availability of multiple distributed points of gas consumption.

Reinforcement learning (RL) is experiencing a resurgence in research interest, where Learning Classifier Systems (LCSs) have been applied for many years. However, traditional Michigan approaches tend to evolve large rule bases that are difficult to interpret or scale to domains beyond standard mazes. A Pittsburgh Genetic Fuzzy System (dubbed Fuzzy MoCoCo) is proposed that utilises both multiobjective and cooperative coevolutionary mechanisms to evolve fuzzy rule-based policies for RL environments. Multiobjectivity in the system is concerned with policy performance vs. complexity. The continuous state RL environment Mountain Car is used as a testing bed for the proposed system. Results show the system is able to effectively explore the trade-off between policy performance and complexity, and learn interpretable, high-performing policies that use as few rules as possible.

This chapter covers three of the most exciting and demanding classes of cloud applications, Big Data, data streaming, and mobile cloud computing. Big Data and data streaming applications require low-latency, scalability, versatility, and a high degree of fault-tolerance. Achieving these qualities at scale is extremely challenging.

Disease dynamics is a complex phenomenon and in order to address these questions expertises from many disciplines need to be integrated. One method that has become particularly important during the past few years is the development of computational models and computer simulations that help addressing these questions. In the focus of this chapter are emergent infectious diseases that bear the potential of spreading across the globe, exemplifying how connectivity in a globalized world has changed the way human-mediated processes evolve in the 21st century. The examples of most successful predictions of disease dynamics given in the chapter illustrate that just feeding better and faster computers with more and more data may not necessarily help understanding the relevant phenomena. It might rather be much more useful to change the conventional way of looking at the patterns and to assume a correspondingly modified viewpoint—as most impressively shown with the examples given in this chapter.

First quantum computers very recently have demonstrated "quantum supremacy" or "quantum advantage": Executing a computation that would have been impossible on a classical machine. Today's quantum computers follow the NISQ paradigm: They exhibit error rates that are much higher than in conventional electronics and have insufficient quantum resources to support powerful error correction protocols. This raises questions which relevant computations are within the reach of NISQ architectures. Several "NISQ-era algorithms" are assumed to match the specifics of such computers; for instance, variational optimisers are based on intertwining relatively short quantum and classical computations, thus maximizing the chances of success. This paper will critically assess the promise and challenge of NISQ computing. What has this field achieved so far, what are we likely to achieve soon, where do we have to be skeptical and wait for the advent of larger-scale fully error-corrected architectures?

Style Transfer has been proposed in a number of fields: fine arts, natural language processing, and fixed trajectories. We scale this concept up to control policies within a Deep Reinforcement Learning infrastructure. Each network is trained to maximize the expected reward, which typically encodes the goal of an action, and can be described as the content. The expressive power of deep neural networks enables encoding a secondary task, which can be described as the style. The Neural Policy Style Transfer (NPST) 1 algorithm is proposed to transfer the style of one policy to another, while maintaining the content of the latter. Different policies are defined via Deep Q-Network architectures. These models are trained using demonstrations through Inverse Reinforcement Learning. Two different sets of user demonstrations are performed, one for content and other for style. Different styles are encoded as defined by user demonstrations. The generated policy is the result of feeding a content policy and a style policy to the NPST algorithm. Experiments are performed in a catch-ball game inspired by the Deep Reinforcement Learning classical Atari games; and a real-world painting scenario with a full-sized humanoid robot, based on previous works of the authors. The implementation of three different Q-Network architectures (Shallow, Deep and Deep Recurrent Q-Network) to encode the policies within the NPST framework is proposed and the results obtained in the experiments with each of these architectures compared.

The chemical production process is a special dynamic and complex system. It has the characteristics of instability and danger, thus making safety management in the production process very difficult. To support the long-term and stable operation of the chemical process, timely and accurate fault diagnosis is very necessary. Aiming at the high-dimensional nonlinearity of chemical process data, this paper proposes a fault diagnosis method in chemical process that is based on multi-scale convolutional neural network (MsCNN) combined with matrix diagram. This model uses software to convert the pre-processed time series signal data of the chemical process into sets of matrix diagrams, and then uses the MsCNN model to accurately diagnose various types of faults in the chemical production process. Taking Tennessee Eastman (TE) process fault data set as an example, verification shows good simulation results in detection accuracy and loss. This method can create high-precision judgments on fault types, which is beneficial to timely elimination of faults and avoidance of safety accidents. The results show that the new model proposed in this paper has good potential in the field of fault diagnosis in chemical processes, and is also reliable, practical and scientific.

The demand for electricity increases rapidly along with the advancement of the industrial age. To ensure efficient distribution of the electricity, maintain low losses and high level of quality, and the security of electricity supply, the smart grid concept was proposed. The concept enables a small, individual scale to generate electricity and sell it to the grid. However, the concept adds complexity to the existing system, such as how a transaction between these generators and consumers are conducted, verified and recorded. This paper proposes the blockchain as a tool to manage transactions in the smart grid. Transactions are performed with smart contracts, and the network acts as a transaction verifier. The blockchain provides immutability of the transactions, which ensure every transaction between generators and consumers will always be executed. It also provides immutability to transaction history, which can be used for audit or solving a transaction dispute.

There is a large growth in hardware and software systems capable of producing vast amounts of image and video data. These systems are rich sources of continuous image and video streams. This motivates researchers to build scalable computer vision systems that utilize data-streaming concepts for processing of visual data streams. However, several challenges exist in building large-scale computer vision systems. For example, computer vision algorithms have different accuracy and speed profiles depending on the content, type, and speed of incoming data. Also, it is not clear how to adaptively tune these algorithms in large-scale systems. These challenges exist because we lack formal frameworks for building and optimizing large-scale visual processing. This paper presents formal methods and algorithms that aim to overcome these challenges and improve building and optimizing large-scale computer vision systems. We describe a formal algebra framework for the mathematical description of computer vision pipelines for processing image and video streams. The algebra naturally describes feedback control and provides a formal and abstract method for optimizing computer vision pipelines. We then show that a general optimizer can be used with the feedback-control mechanisms of our stream algebra to provide a common online parameter optimization method for computer vision pipelines.

Conventional optimization algorithms that prescribe order packing instructions (which items to pack in which sequence in which box) focus on box volume utilization yet tend to overlook human behavioral deviations. We observe that packing workers at the warehouses of the Alibaba Group deviate from algorithmic prescriptions for 5.8% of packages, and these deviations increase packing time and reduce operational efficiency. We posit two mechanisms and demonstrate that they result in two types of deviations: (1) information deviations stem from workers having more information and in turn better solutions than the algorithm; and (2) complexity deviations result from workers’ aversion, inability, or discretion to precisely implement algorithmic prescriptions. We propose a new “human-centric bin packing algorithm” that anticipates and incorporates human deviations to reduce deviations and improve performance. It predicts when workers are more likely to switch to larger boxes using machine learning techniques and then proactively adjusts the algorithmic prescriptions of those “targeted packages.” We conducted a large-scale randomized field experiment with the Alibaba Group. Orders were randomly assigned to either the new algorithm (treatment group) or Alibaba’s original algorithm (control group). Our field experiment results show that our new algorithm lowers the rate of switching to larger boxes from 29.5% to 23.8% for targeted packages and reduces the average packing time of targeted packages by 4.5%. This idea of incorporating human deviations to improve optimization algorithms could also be generalized to other processes in logistics and operations. This paper was accepted by Charles Corbett, operations management.

Visual Question Answering (VQA) has attracted extensive research focus recently. Along with the ever-increasing data scale and model complexity, the enormous training cost has become an emerging challenge for VQA. In this article, we show such a massive training cost is indeed plague. In contrast, a fine-grained design of the learning paradigm can be extremely beneficial in terms of both training efficiency and model accuracy. In particular, we argue that there exist two essential and unexplored issues in the existing VQA training paradigm that randomly samples data in each epoch, namely, the "difficulty diversity" and the "label redundancy". Concretely, "difficulty diversity" refers to the varying difficulty levels of different question types, while "label redundancy" refers to the redundant and noisy labels contained in individual question type. To tackle these two issues, in this article we propose a fine-grained VQA learning paradigm with an actor-critic based learning agent, termed FG-A1C. Instead of using all training data from scratch, FG-A1C includes a learning agent that adaptively and intelligently schedules the most difficult question types in each training epoch. Subsequently, two curriculum learning based schemes are further designed to identify the most useful data to be learned within each inidividual question type. We conduct extensive experiments on the VQA2.0 and VQA-CP v2 datasets, which demonstrate the significant benefits of our approach. For instance, on VQA-CP v2, with less than 75 percent of the training data, our learning paradigms can help the model achieves better performance than using the whole dataset. Meanwhile, we also shows the effectivenesss of our method in guiding data labeling. Finally, the proposed paradigm can be seamlessly integrated with any cutting-edge VQA models, without modifying their structures.

Governments and industries continuously engage in large-scale projects aiming to unlock economic potential for commercial and social benefits of a country and its people. In the case of developing countries, where the dynamics of innovation ecosystems differ from those in developed countries, the need for sociotechnical transition projects are often quite high. The aim of this article is to propose a standardized framework suitable for developing countries for the technology roadmapping of large projects aimed to bring about systemic sociotechnical transitions. The framework is developed through the integration of existing generic technology roadmapping approaches with transition management theory and complex systems theory and was illustrated against a successful sociotechnical transition program that is characterized by Big Fast Results (BFR) requirements. Data were collected from qualitative interviews as well as from the critical analysis of relevant documents. The proposed sociotechnical transition technology roadmap accurately captured the key elements of the BFR project it was benchmarked against and even identified key elements that were not included in the BFR methodology. An advantage of the proposed sociotechnical transition technology framework is the fact that it makes use of a standard technology roadmapping process with few customizations that are ideal for use in developing countries.

The theory of capacities provides powerful formal methodology to account for criteria dependencies in multiple criteria decision problems. The discrete Choquet and Sugeno integrals aggregate criteria valuations accounting for criteria synergies and redundancies. We address an important problem of randomly generating capacities of special classes for simulation studies and for capacity learning through evolutionary algorithms. We discuss two efficient methods suitable for k-interactive capacities. The results are supported by the extensive numerical evidence and provide a useful tool for large scale simulations.

Solid phase extraction is widely used in sample pretreatment, concentration and analysis processes due to high selectivity and suitability for low concentration sample system. In this review, we systematically summarized and discussed the development trends of solid phase extraction by bibliometrics method. By analyzing papers output scale, the research and development direction of solid phase extraction technology is prospected. We also give an overview on current strategies of novel solid phase extraction from the separation medium and separation technology. The paper aims to describe the global research profile and the development trends of solid phase extraction, to help researchers to accurately grasp the research trend and to provide support for scientific research institutions to formulate scientific policies and strategic plans. Furthermore, the prospect of the development and application of solid phase extraction is also discussed.

Background: The prediction of drug-protein interaction (DPI) plays an important role in drug discovery and repositioning. Unfortunately, traditional experimental validation of DPIs is expensive and time-consuming. Therefore, it is necessary to develop in silico methods for the identification of potential DPIs. Method: In this work, the identification of DPIs was performed by the generated recommendation of the unexplored interaction of the drug-protein bipartite graph. Three kinds of recommenders were proposed to predict the potential DPIs. Results: The simulation results showed that the proposed models obtained good performance in crossvalidation and independent test. Conclusion: Our recommendation strategy based on collaborative filtering can effectively improve the DPI identification performance, especially for certain DPIs lacking chemical structure similarity or genomic sequence similarity.

This work explores additive manufacturing (AM) of concrete by using a six-axis robotic arm and its use in large-scale, autonomous concrete construction. Concrete AM uses an extrusion method to deposit concrete beads in layers to create a three-dimensional (3D) shape. This method has been found to have many uses and advantages in construction applications. The lack of formwork and autonomous nature of this manufacturing method allows for new geometries and materials to be printed in unsafe or challenging environments. Autonomous construction has been suggested as a method of creating habitats in rapid-response scenarios. This article discusses research toward one such system that could be used to rapidly construct necessary habitats in response to low-resource and emergency situations. This required addressing certain limitations of a six-axis robotic arm platform along with overcoming system challenges to achieve deliverables for NASA's “3D Printed Habitat Challenge.” This included system design to increase the build volume, integrate embedding, print non-coplanar sections, and minimize travel moves to address the challenges associated with continuous extrusion of cementitious material. The system was demonstrated by printing a one-third scale habitat, which represents the first 3d-printed fully enclosed structure at an architectural scale without the use of support.

The issue of cutting fluid is of great concern in present scenario of industry 4.0 due to its application in machining operation performed at wider parameter ranges. Despite of numerous advantages and enhanced machining performance, few critical issues like environmental and health hazards occurred if not controlled in efficient manner, consequently, set negative impacts on their utility. Hence, to safeguard aforementioned problems and to achieve environmental sustainability, the bio-degradable oils are gaining popularity due to their useful impact attained through latest lubrication techniques and amalgamation of nano and ionic fluids. Further, the cost of later fluids forms a barrier to their application in present competitive industrial framework as compared to conventional cutting fluids. In present investigation the survey was conducted on the scenario of cutting fluids application in small scale industries present in the Malwa belt of Punjab. The research was performed through well designed questionnaire consisting of various questions related to different aspects of cutting fluids such as fluid type, volume, application methods, recycling approach, disposal way and most important health as well as environmental issues. The data was collected by qualified engineers from various industries and further confirmed from the official source for its validity prior to analysis. The results of investigation confirmed that there was deficiency of advanced lubrication techniques utilization, insufficiency of post treatment and few cases of health hazards.

Innovations have the potential to help us address and overcome many of the challenges that agriculture is facing today. Yet, at the same time, they have the potential to create new, sometimes even more challenging, problems, especially when they are not governed in a sustainable way. Governing agricultural innovation sustainably requires understanding of all components that influence, and are influenced by, innovation processes, interactions across societal levels, and the normative and power dynamics that come together to shape the direction and outcomes of innovation processes. Hitherto, approaches to (agricultural) innovation and transition tend to specialize on a specific societal scale or sub-aspect of innovation or transition processes. In this article we aim to bring the strengths of some of the main approaches (Multi-Level Perspective, Agricultural Innovation Systems, Responsible Innovation, Innovation Management, Theory of Planned Behaviour) and insights from environmental governance literature together into a comprehensive framework. The framework describes seven key components and their interactions: macro context, governance system, immediate context, innovative and adaptive capacity of the actors, psychosocial factors, and the innovation process itself. Based on these, we present a subset of guiding questions that can be used diagnostically or for design purposes to support the sustainable governance of agricultural innovation processes.

Remote management is a modular and effective approach that allows for remote control and monitoring of irrigation systems and hydro-mechanical facilities that are spatially separated, while also tracking real-time irrigation data such as flow, pressure, electric pump groups, solar panels, and agro-meteorological data. Control instrumentation and automation equipment indicate the various defects that may occur on the hydro-mechanical equipment, which can upset the normal functioning of the structures of the irrigation system. Identifying the level of operation of the remote management system is an essential step to understand better the functionality of the remote control and monitoring system and define the objectives to be met and the technical specifications necessary for its implementation. This article will present a methodological approach for the implementation of a remote management system for large-scale irrigation. The proposed methodological approach focuses on improving the hydraulic parameters of pumping stations and irrigation hydrants and decreasing pumping station electricity consumption.

This study proposes the integrated designs of energy systems and a re-liquefaction system for ocean-going LH2 tankers. Five prospective energy systems (Systems A to E) are suggested, using LNG as fuel, and a re-liquefaction system with a Claude cycle is developed. Their economic value, technological feasibility, and environmental impact are evaluated. The re-liquefaction systems' exergy efficiency and specific energy consumption ranges were 26.79–46.27% and 3–7.45 kWh/kg, respectively. The re-liquefaction system in System C is economically feasible up to $2/kg of H2. LCC of the integrated designs shows that System C has the lowest cost of $140 million. The shipping costs for each design are reviewed, and the lowest one is $447/ton of H2 for System C. Although System C CAPEX is the second expensive, it has the highest efficiency. Consequently, System C with the low-pressure engine, SOFC, and the re-liquefaction system is determined to be the optimal one.

Thermal energy storage systems utilising phase change materials have the potential to overcome the intermittency issues associated with most renewable energy sources, significantly contributing to the decarbonisation of the energy sector. While the concept of storing energy in the latent heat of a phase transition is not new, large scale systems employing this concept have not yet realised their full potential. This is in large part due to the shortcomings of conventionally used materials – whether it be issues with their flammability, chemical and thermal instability, corrosivity, poor cycling stability, or a combination of these. Ionic liquids offer a suite of inherent “green” properties that translate well into the field of phase change materials, namely low volatility, low flammability, and good thermal and chemical stability, alongside the potential to tune their chemical and phase properties. In this Perspective, we discuss the evolution and promise of the emerging field of ionic liquids for renewable thermal energy storage. Systems are considered from a holistic, sustainable point of view, demonstrating the importance of assessing material origins and synthetic pathways as well as system performance through lifecycle assessment. We elucidate the emerging design rules for optimising thermal properties, and in doing so attempt to provide an overview of promising emerging systems and future directions.

Graphical abstract, Highlights • A comprehensive analysis of modern object detection models • Study on models that can be used as detectors for defect detection applications in industry. • Study on YOLOv4 that can perform good defect detection with not much capital investment. • Analysis on the correlation between dataset, labeling and the data augmentation steps and accuracy and computations. • Analysis on the importance of correct data acquiring, augmentation and labeling in low cost applications. • Analysis of the rate of improvement of the mAP of defect detection and image classification systems in recent years. • Analysis of model compression and acceleration on embedded applications and smart factories., Background Over the last years Deep Learning has shown to yield remarkable results when compared to traditional computer vision algorithms, in a large variety of computer vision applications. The deeplearning models outperformed in both accuracy and processing time. Thus, once a deeplearning models won the Image Net Large Scale Visual Recognition Contest, it proved that this area of research is of great potential. Furthermore, these increases in recognition performance resulted in more applied research and thus, more applications where deeplearning is useful: one of which is defect detection (or visual defect detection). In the last few years, deeplearning models achieved higher and higher accuracy on the complex testing datasets used for benchmarking. This surge in accuracy and usage is also supported (besides swarms of researchers pouring into the race), by incremental breakthroughs in computing hardware: such as more powerful GPUs(Graphical processing units), CPUs(central processing units) and better computing procedures (libraries and frameworks). Aim of the review To offer a structured and analytical overview(stating both advantages and disadvantages) of the existing popular object detection models that can be re-purposed for defect detection: such as Region based CNNs(Convolutional neural networks), YOLO(You only look once), SSD(single shot detectors) and cascaded architectures. A further brief summary on model compression and acceleration techniques that enabled the portability of deeplearning detection models is included. Key Scientific Concepts of Review It is of great use for future developments in the manufacturing industry that many of the popular, above mentioned models are easy to re-purpose for defect detection and, thus could really contribute to the overall increase in productivity of this sector. Moreover, in the experiment performed the YOLOv4 model was trained and re-purposed for industrial cable detection in several hours. The computing needs could be fulfilled by a general purpose computer or by a high-performance desktop setup, depending on the specificity of the application. Hence, the barrier of computing shall be somewhat easier to climb for all types of businesses.

Two-dimensional (2D) crystals of conjugated polymers (2DCCPs) have attracted significant attention in the past few decades due to their superior and intriguing optoelectronic properties. With the increase in brain-inspired neuromorphic computing, 2DCCPs have also been employed in the construction of synaptic devices because they offer clear advantages of easy integration, high stability, long–range order and no biological rejection. However, pursuing controllable strategies to obtain high-quality and large-scale 2DCCPs remains a challenge on account of their intrinsic complex structures. Herein, we present a universal methodology to grow centimeter-scale 2DCCPs in a controlled manner within minutes by selecting an appropriate growth-assisting solvent and surfactant. The thickness of the as-produced 2DCCPs was down to two molecular layers. The photonic synaptic transistors based on the as-grown 2DCCPs could perform typical synaptic functions, including short-term plasticity, paired-pulse facilitation, long-term plasticity and spike intensity-dependent plasticity. The transistor was also proven to have potential application as a high-pass filter. This methodology shows great potential to be expanded to many other polymer systems and opens new pathways for large-area integrated wireless communication systems and artificial visual systems.

The manufacturing world had undergone four separate industrial revolts. Industry 4.0 is a most recentchapter in the Industrial revolt story that pivotedmainly on manufacturing automation, live data capturing, smart machines, and internet connectivity with key objectivesof productivity rise, optimization of available working hours, and increased organization competitiveness. This research paper studiesthe industry 4.0 through the survey of available literature in a sequential manner that started with understanding the development of the manufacturing industry from industry 1.0 to industry 4.0, the nine key foundation blocks of the industry 4.0, the cyber-physical systems, the Smart factory concept,industry 4.0 application strategy guide step by step, Industry 4.0 benefits, risk, and challenges. We applied the Industry 4.0 concept step by step as a pilot project as per the Industry 4.0 application strategy guide atthe bus bar trunking production factory with the target of delivering the demand of customerswithin committed or shorter lead time from the bus bar manufacturing shop and discussed the outcomes briefly. We concluded that Industry 4.0 helps to improve operator work-life, overall productivity, and to reduce the operating cost of production drastically in mass manufacturing processes. Industry 4.0 is also a boon in today’s changing work culture after the CORONA pandemic through its core concepts of digitization, paperless work, networking, and maximum internet usage. It also guides to adopt automation and smart machines that are communicating in real-time with its customer. We also observed that its application is costly in terms of investment to deploy for small or middle scale manufacturing plants, but they shall implement partly to get the benefit from it.

Material Extrusion Additive Manufacturing (MEAM) method has been focused for its capability to fabricate complex geometrical parts without any specific tooling and its suitability for customized parts production. Many studies have recognized the role of MEAM methods in small scale manufacturing and in-house fabrication. But higher surface roughness in material extrusion’ parts is a major drawback which limits its industrial applications. The obtained surface roughness of fabricated parts beyond the desired limits needs further investigations. Since the surface roughness in the parts can't be totally eliminated but it can be reduced significantly. Various researchers have quantified the effect of extrusion parameters and some of them have worked on the post processing methods. This literature work focuses on the effect of majorly used extrusion variables and post processing techniques for the improvements of surface quality.

This article develops a system dynamics model for innovation ecosystems in the high-tech industry. The model emphasizes the interactions between high-tech enterprises, the government, and universities in the innovation process. In particular, we use two high-tech instruments, the Mars Rock Corer and the Soil Preparation System, in the aerospace innovations developed by Hong Kong Polytechnic University. These instruments are a byproduct of collaborations with the European Space Agency, Russian Space Agency, and China Space Agency. They span decades of space missions to the moon and Mars. Furthermore, we propose a three parities systems dynamics model to study innovation behaviors. This model can be used to study and analyze the stakeholders’ interactions in high-technology-oriented innovation ecosystems. Theoretical, computational, and simulation studies are also conducted to understand and evaluate the evolution of the ecosystem. The results demonstrate the feasibility and effectiveness of this approach, which can be applied in on a large scale and in complex space projects.