ABSTRACT. In linear temporal logic (LTL) model checking using Petri nets, two important aspects are state generation and exploration for counterexample search. Traditional state generation involves updating a structure of enabled transitions and frequently encoding/decoding to read each encoded state, which can be expensive. This paper proposes an optimized approach that calculates enabled transitions on demand using a dynamic fireset, avoiding the need for such a structure. Additionally, a set of direct read/write (DRW) operations on encoded markings is proposed to speed up state generation and reduce memory usage without the need for decoding and re-encoding. Heuristic information is incorporated into the Büchi automaton for counterexample search to guide exploration toward accepted states. These strategies improve existing methods for LTL model checking with Petri nets. The optimization strategies are implemented in a tool called EnPAC (Enhanced Petri-net Analyser and Checker) for linear temporal logic and evaluated on MCC (Model Checking Contest) benchmarks, demonstrating a significant improvement over previous methods.

ABSTRACT. We develop a tool named PNer, a model checker for verifying Computation Tree Logic based on Petri nets. This tool includes three parts. The first one can verify concurrent systems that are modeled by Petri nets for their properties/requirements specified by Computation Tree Logic. The second one can verify multi-agent systems that are modeled by knowledge-oriented Petri nets for their properties/requirements specified by Computation Tree Logic of Knowledge. The last one can verify real-time systems that are modeled by prioritized time-point-interval Petri nets for their properties/requirements specified by Timed Computation Tree Logic. For the first and second ones, PNer supports efficient symbolic techniques based on Ordered Binary Decision Diagrams to alleviate the state explosion problem. For the last one, PNer can quantitatively analyse time-related requirements of a real-time system, i.e., computers the minimum or maximum time bound ensuring a real-time system satisfies such a time-related requirement. In this paper we introduce PNer.

ABSTRACT. Cluster tools are widely used for wafer fabrication in semiconductor fabs. For some wafer fabrication processes, cluster tools should deal with reentrant wafer flows. In such cases, it is necessary to find a way to correctly operate cluster tools such that wafers can be completed according to wafer processing recipe (including the processing parameters and wafer flows) and the productivity of the tools can be maximized. To do so, this work presents a Petri net model for dual-arm cluster tools with multiple-time reentrant flows. Then, based on the Petri net model, control policies are proposed such that several novel scheduling methods can be realize in practice. Note that such control policies can be embedded into the scheduler or controller of cluster tools. Finally, case studies are given to show how the control policies are used to operate cluster tools.

ABSTRACT. Cloud Computing has emerged as a prominent paradigm for utilizing on-demand computing, storage, and network services in an efficient manner. With the increasing demand for cloud resources, task scheduling and data placement have become a significant research topic in this domain. An efficient task scheduling aims to assign tasks to virtual machines in order to minimize workflow execution time. However, the scheduling of workflows is known to be an NP- complete problem, posing computational challenges. This paper presents an adaptation of bio-inspired method called Discrete Symbiotic Organism (DSOS) in the field of scientific workflow scheduling, The aim of this proposed algorithm is to minimize the completion time and transfer time of data. The performance of this proposed algorithm has been evaluated using the CloudSim toolkit. Simulation results show the effectiveness of the proposed algorithm in comparison with well-known algorithms.

ABSTRACT. In recent times, there has been a notable shift in manual assembly tasks being replaced by robots. These robots are controlled by operators using keyboards. To enhance this practice, we have introduced a remote human-robot collaboration system. This system allows operators to remotely control robots using hand gestures, utilizing an online hand gesture recognition system. To complement the remote robot control system, a model-driven display system has been implemented.

For data acquisition, we have carefully selected Leap Motion technology. The designed approach for remote hand gesture recognition has been implemented and tested for accuracy and response time using the well-known SHREC'21 Gesture Benchmark. The final analysis demonstrates the significant potential of the developed system in improving assembly processes and environments. It optimizes various key performance indicators, leading to enhanced efficiency and effectiveness.

ABSTRACT. For better quality control and reducing the system load, twin-cluster tools are widely used in recent years. Due to the different construction of the tool compared to the traditional multi-cluster tool, it is difficult to schedule the twin-cluster tool with existing studies. This paper aims to find a feasible periodic schedule method for the twin-cluster tool. First, the workload of each processing chamber in the tool are analyzed such that the optimal cycle time can be obtained. Then, a mathematical programming model is presented to achieve the minimal cycle time. A novel algorithm is established so as to obtain the robot waiting time. Then, several sufficient conditions that ensures the twin-cluster tool achieves the optimal cycle time are proposed. Finally, numerous experiments are conducted to verify the accuracy of the conclusion with different conditions.

ABSTRACT. This paper examines the application of modelfree control as introduced by Fliess and Join, to tactical inventory control in a supply chain system. The presence of the unavoidable delay which is inherent to such system requires a compensation action. Here, the delay compensation is achieved via a forecasting method stemming from the new advances of time series. The conducted simulations and comparative studies of a real industrial example show improved performance, greater flexibility, and higher functionality of the proposed framework even in the presence of various disturbances and sever constraints.

ABSTRACT. Customer demands for personalized products and fast delivery have become increasingly significant in e-commerce competition. Additive manufacturing (AM) has emerged as a solution to address the requirements of customized production. A recent advancement in AM, the mobile mini-factory, enables a truck equipped with a single 3D printer to produce orders en-route to customer's location. This innovative mode of simultaneous production and transportation can reduce delivery times and storage expenses for companies. In this study, we investigate a simultaneous production and transportation problem to minimize the total cost of waiting time and tardiness (referred to as SPTP-CWT). The problem is formulated as a mixed integer linear programming (MILP) model. A heuristic approach, simulated annealing with hybrid mutation algorithm (SA-HM), is developed to solve large-sized instances. Computational experiments are conducted on benchmark instances and the results indicate that the proposed SA-HM is capable to give better solutions within a reasonable computation time when compared to directly solving the MILP model and an ant colony optimization algorithm.

ABSTRACT. Timely and accurate credit card fraud detection (CCFD) is concerned by all financial institutions. Existing CCFD methods generally employ aggregated or raw features as their representations to train their detection models. Yet such features tend to fall short of effectively exposing some critical characteristics of frauds. In this work, we propose a spatial-temporal gated network (STGN) to automatically learn new informative transactional representations containing users' transactional behavioral information for CCFD. A special gated recurrent neural net unit is constructed with a time-aware gate and location-aware gate to extract users' spatial and temporal transactional behaviors. A spatial-temporal attention module is designed to expose the transaction motive of users in their historical transactional behaviors, which allows the proposed model to better extract the fraudulent characteristics from successive transactions with time and location information. A representation interaction module is offered to make rational decisions and learn compositive transactional representations. A real-world transaction dataset is used in our experiments to verify the efficacy of the learned new representations. The results demonstrate that our proposed model outperforms the state-of-the-art ones, thus greatly advancing the field of CCFD.

ABSTRACT. Nowadays, online bullying, called cyberbullying, has become one of the most common problems faced by young people in society. The text, containing and spreading amounts of information, is a key medium in the social media and social networks. The natural language processing (NLP) method as the most common method deals with textual information. In the field of cyberbullying detection, extracting information from texts is a way to predict online bullying phenomenon presently. In this work, we propose a transformer based hierarchical neural network (THNN). Compared to existing methods, ours has the following advantages: (i) it captures longer context information for longer time intervals; (ii) it reflects and mimics the hierarchical structure of conversation media clearly; (iii) it has less computational time.

ABSTRACT. As a highly effective tool for enterprise strategic planning and technology management, the technology roadmap (TRM) has garnered increasing attention from enterprises. This paper aims to address the issue of evaluating and making decisions on technical solutions during the process of drawing an initial TRM using traditional drawing techniques, such as T-plans. After conducting in-depth research on the TRM drawing process, we propose introducing wargame deduction into the process of drawing an enterprise's TRM, thereby expanding the process of drawing a roadmap and enriching the theory behind the TRM drawing process. We have designed a set of related wargames for the evaluation of plans during the production process of enterprise TRMs. Finally, we present an analysis of the advantages and disadvantages of using wargames to draw TRMs and conclude that wargame deduction is a feasible approach to achieve intelligent TRM drawing.

ABSTRACT. This paper delves into the nascent paradigm of Explainable AI (XAI) and its pivotal role in enhancing the acceptability of growing AI-systems that are shaping the Digital Management 5.0 era. XAI holds great promise, promoting compliance with legal and ethical standards, and offering transparent decision-making tools. The imperative of interpretable AI systems to counter the black box effect and adhere to data protection laws like GDPR is highlighted. This paper aims to achieve a dual objective. Firstly, it provides an in-depth understanding of the emerging XAI paradigm, helping practitioners and academics project their future research trajectory. Secondly, it proposes a new taxonomy of XAI models with a collection of potential applications that could facilitate AI acceptability. Although the academic literature reflects a crucial lack of exploration into the full potential of XAI, existing models remain mainly theoretical and lack practical applications. By bridging the gap between abstract models and the pragmatic implementation of XAI in management, this paper breaks new ground by launching the scientific foundations of XAI in the upcoming era of Digital Management 5.0.

ABSTRACT. To address the limitations of conventional offshore oil-and-gas field inspections, such as high cost, low frequency, and high risk, a new method called vessel-drone collaborative inspection is proposed. This approach involves a drone-carrying vessel departing from the port and sailing to a designated distance from the oil-and-gas field. The drones are then released to conduct field inspections. Once the drones complete their tasks, the vessel retrieves them at a specific location. This paper focuses on addressing the multi-period "vessel-drone" collaborative inspection route planning problem. The objective is to minimize the fixed cost of the vessel over the planning horizon while considering the overall path cost of both the vessel and the drone. The planning model takes into account various factors, including the coordination of space and time between the vessel and drone, drone endurance limitations, and the inspection cycle limitations for each period. To solve the problem, a mixed-integer second-order cone optimization formulation is proposed, along with several valid inequalities. Through empirical research and sensitivity analyses, the effectiveness of the proposed method is demonstrated. It is shown that the method can provide a cost-effective plan for oilfield inspections, offering potential value in guiding offshore oil-and-gas field inspection operations.

ABSTRACT. This paper discusses the use of the cell transmission model to evaluate the performance of urban traffic networks, specifically in addressing incident-based traffic congestion. The authors present an extended traffic control policy utilizing the cell transmission model to simulate and analyze urban traffic jam problems. We develop a traffic simulation tool based on the cell transmission model using cross-platform, which effectively models the propagation and dispersion of traffic congestion resulting from an accident. The study investigates the application of the cell transmission model to develop control strategies for dispersing accident-induced traffic jams and assesses the effectiveness of these strategies. Overall, this work demonstrates the potential of the cell transmission model as a valuable tool for managing traffic flow in urban areas.

ABSTRACT. In the context of vehicle task offloading, this research introduces a novel approach to address the challenge of insufficient efficiency in multi-task offloading within a cloud-edge hybrid environment. The proposed method leverages a probabilistic performance-aware evolutionary game strategy. To tackle the varying edge server performance in a heterogeneous setting comprising central and edge servers, historical performance data undergoes probabilistic analysis. This leads to the construction of an evolutionary game model. Subsequently, an evolutionary stable strategy (ESS) is devised for service offloading, ensuring high user satisfaction. Through simulation experiments across 24 consecutive time windows using real-world datasets, our framework demonstrates superior performance across various metrics when compared to existing methods.

ABSTRACT. This work proposes the application of reinforce- ment learning approaches to vehicle automatic control. It addresses yaw motion stability and lane-keeping maneuvers. Reinforcement learning is used to tune PID parameters using LQG cost function as a reward function to optimize the PID parameters. The cost combines tracking error reduction and actuator effort limitation. Two learning agent performances are explored: the simple deep deterministic policy gradient (DDPG), and a more sophisticated one called Twin Delayed Deep Deterministic (TD3). The learning procedure checks the closed-loop stability before parameters update. Both DDPG and TD3 are compared in terms of performance to the Matlab- Simulink control parameters tuner. Yaw rate profile tracking and lane-keeping maneuvers are used to compare the behaviour of the achieved controllers. TD3 is found to be able to provide faster response while ensuring sufficient phase margin.

ABSTRACT. Opacity is a confidentiality property that holds if certain secret behavior of a system cannot be revealed under any system evolution. One way to enforce opacity is by using an extended insertion function, which is capable of inserting fake symbols before and after an actual system output in real time as the system evolves. This paper studies the enforcement of opacity for systems modeled by finite state automata using extended insertion functions under event insertion constraints and limited insertion event constraints, i.e., only a limited number of symbols in a specified set can be inserted before and after an actual system output. The notion of $(k_b,k_a)$-EIC-enforceability is proposed to describe how to enforce opacity via extended insertion functions that are restricted in this way. We also construct a verifier that can be used to check a necessary and sufficient condition for opacity enforceability.

ABSTRACT. This paper addresses the performance evaluation of fault detection and diagnosis (FDD) functions included in most SCADA systems. Our aim is to enlarge the discrete event system perspective about fault detection and diagnosis. In fact most of the existing approaches are only interested in the possible non detection situations, but are designed such that they do not provide any false alarm. This situation may lead to poor performance regarding the non detection rate. Here a systematic schema is proposed to study both reactivity (non detection) and correctness (false alarm) properties. The approach relies on the Markovian approximation of the system behaviour in the long run and the proposed schema is developed according to two successive product compositions: at a first time with the fault model, and a second time with the FDD function. As a result, a probabilistic verifier that explains the main average indicators to evaluate the performance of the SCADA system, is presented. The approach is suitable for single fault event as for more complicated fault patterns.

ABSTRACT. This paper deals with discrete event systems modeled by output synchronized Petri nets. This formalism allows to represent the system behavior along with the received control inputs and sensors outputs that circulate through a communication network. We suppose that the system includes multiple deadlocks and is vulnerable to cyber-attacks that aim to drive this system to one of these deadlocks. We assume that the attacker is able to insert control inputs and to read/delete the sensors outputs. Moreover, the current state of the system is supposed to be unknown by the attacker. In this case, a synchronizing sequence, which is a set of input control events, can be used to drive the system to a known state. In this work, an attacker's perspective is adopted and a method to compute an attack Synchronizing Sequence (attack SS) is proposed. To do so, all deadlocks are merged into a single one and attack SS that drive the system state to one of its deadlocks without knowing the actual state of the system are computed. For each computed SS, the associated sensors outputs are generated. This allows the attacker to choose the best synchronizing sequence.

ABSTRACT. In this paper, we examine the supervisory control problem for cyber-physical systems vulnerable to actuator jamming attacks, which cause control command loss during transmission. We study this issue within the framework of discrete event systems. First, an attack structure is constructed by analyzing the closed-loop control behavior under jamming attacks. It shows how intruders employ jamming attacks to drive a plant into unsafe states. We then explore a solution to the supervisory control problem by introducing two concepts, i.e., dynamically controllable languages and free states. In particular, an existence condition for a nonblocking and robust supervisor is obtained, and an algorithm deriving a supervisor capable of defending against actuator jamming attacks is proposed.

ABSTRACT. Execution ability is a manifestation of system efficiency, and expectations are the goals pursued by the system. However, the existing \emph{Command and Control} (C2) model does not fully describe and reflect this. Therefore, by integrating the advantages of classical command and control models, expected utility theory and management theory, a command and control model that considers execution ability and expectation is proposed —— the Perception-Reasoning-Evaluation-Decision-Response-Operation-Storage-Expectation (PREDROSE) cycle. This paper provides a detailed explanation of the response stage for evaluating execution ability and the expectation module for calculating expectations in the model, and discusses various techniques and methods involved.

ABSTRACT. The task allocation between team members of each organization's departments is a crucial project management activity that needs adequate attention to improve organizational efficiency. If a related decision is not made properly, it may immediately cause problems with coordination and communication, task delays, and increased costs. Tasks are usually grouped and then these groups are assigned to the agents. Moreover, it is important to consider the difficulty level of the tasks and the member’s ability. The task allocation process becomes increasingly complex when administrators strive to maintain the balance between the workload of allocated task groups and the agents' capabilities. The E-CARGO (Environments - Classes, Agents, Roles, Groups, and Objects) model, created for the purpose of Role-Based Collaboration (RBC), has been effectively utilized to address the given problem. In this paper, a novel approach is presented to formalize the Grouped Task Allocation Problem (GTAP) employing an extended Group Multirole Assignment with Conflicting Roles problem (GMRACR), which is a sub-model of E-CARGO. In this paper, a role negotiation method is introduced, which is based on GMRACR. It employs a partitioning clustering algorithm along with the Analytic Hierarchy Process (AHP) to evaluate the difficulty level of tasks and the abilities of agents, respectively. This formalization of GTAP facilitates the identification of a solution using the IBM ILOG CPLEX optimization package (CPLEX). The proposed method helps decision-makers with a framework to successfully assign balanced grouped tasks. Through simulation experiments on a real-world problem, the efficacy of the suggested approach is substantiated. Experimental results reveal the practicality of the solutions recommended in this paper.

ABSTRACT. In the assembly of the engine rotor, the unbalance quantity of the rotor has a profound impact on equipment quality. Specifically, the selection and position of blades are crucial factors that determine the level of rotor unbalance. Aiming at the multiple installations and repeated adjustments in the assembly process, we study a variant of the single-stage blade selection and position problem (SSBSP) by considering the tolerance criteria. In this paper, we formulate a novel mixed-integer programming model that extends to multiple-rotor assembly for the first time. The objective is to minimize the maximum residual unbalance. Due to the NP-hard in the strong sense of engine rotor blade selection and position, we devise an improved partheno-genetic algorithm to access a near-optimal solution. Furthermore, local search is introduced for deep search and aims to improve the efficiency of the algorithm. A computational analysis is carried out with the aim of analyzing the performance of our methodology in terms of effectiveness and efficiency. Our experimental results demonstrate the superior performance of the proposed algorithm when compared against state-of-the-art heuristics in the literature in terms of solution quality and efficiency.

ABSTRACT. In laser powder bed fusion additive manufacturing, using fixed values for the process parameters (beam power, velocity, and other parameters) may not lead to homogeneously distributed heat in all locations in the build, especially around complex design features. This could lead to builds with defects, leading to poor mechanical and micro-structure properties. To guarantee heat homogeneity, the process parameters need to be actively controlled to adapt to different locations in the build. Builds with varying geometrical features would need different control strategies. In this work, we propose to use reinforcement learning (RL) to control, for the first time, simultaneously multiple AM process parameters to achieve consistent melting properties. Our results show that using RL as a multiple-input multiple-output control system achieves a more consistent meltpool geometry.

ABSTRACT. For the application of industrial robots, positioning accuracy is a significant indicator of their performance. Aiming at the issues of low positioning accuracy, the calibration techniques are employed to obtain the real kinematic parameters for effectively enhancing its accuracy. However, in practical scenarios, for the restriction of the robot workspace, the samples collected during the robot calibration process cannot cover the entire space of the entire space, resulting in an impact on data integrity. To address the above issues, we develop a calibrator integrating the MSS method (measurement samples selection) and LM algorithm (Levenberg-Marquardt) with a spatial line constraint (LMLC), which contains three-fold: a) selecting a set of most representative measurement samples according to the observability index for enhancing the stability of calibration results; b) develop an LM algorithm with line constraint to solve the problem of spatial restriction of robot sampling; c) presenting a robot calibrator that combines MSS and LMLC for efficiently improving the robot calibration accuracy. Experiments illustrate that the MSS-LMLC calibrator outperforms state-of-the-art calibrators on an industrial robot’s calibration accuracy and computational efficiency.

ABSTRACT. This paper deals with the modelling of a new energy efficient scheduling problem. More specifically, it focuses on a multi-line hybrid flow shop. The problem consists of optimizing total energy cost under Time-of-Use pricing with respect to additional constraints: (i) total energy consumption, (ii) peak power limitations and (iii) makespan. An exact solving approach is investigated and results are analysed on a sample of randomly generated instances.

ABSTRACT. Abstract—In many developing countries, handwritten medical prescriptions remain a prevalent mode of documentation within healthcare systems. However, the legibility of these prescriptions often poses a significant challenge, leading to potential errors in patient care. To address this critical issue, our work focuses on developing a model for the recognition of medical hand- written prescriptions. The model architecture is composed of Convolutional Neural Networks (CNNs) and Bi-directional Long- short term memory (Bi-LSTMs) and a Connectionist temporal classification (CTC) Layer to compute loss between predicted sequences and ground truth sequences. However, the availability of annotated medical handwriting data is constrained, particu- larly due to privacy concerns. In response, we delve into transfer learning techniques to harness existing knowledge and enhance model performance. By leveraging transfer learning, we aim to optimize our recognition model using fine-tuning strategies. This study contributes to the field of medical handwriting recognition by exploring the potential of transfer learning in enhancing model accuracy and applicability within developing healthcare contexts.

ABSTRACT. The error suppression and compensation of conventional ultra-wideband (UWB) ranging are less effective in non-line-of-sight (NLOS) environments. The contribution of obstacles that cause NLOS range errors to environmental perception is neglected. In order to achieve a comprehensive perception and understanding of the environment under dynamic positioning scenarios, this work proposes a UWB-based obstacle recognition method. Firstly, the channel impulse response (CIR) curve and communication bottom layer characteristics of UWB in a communication process are analyzed. The key parameters of signal features are extracted and combined with the integrated learning XGBoost classifier. Further, an efficient NLOS obstacle recognition method based on a loss function probability matrix weighted prediction label is proposed. The predicted probabilities of the XGBoost loss function are used as the weights of the labels, and the UWB data of continuous time series is weighted average. This algorithm mitigates the effect of low probability data on the overall prediction results. It also helps to improve the obstacle recognition accuracy. Finally, experiments under dynamic and static conditions are carried out to verify the reliability of the proposed method. The experimental results show that the recognition accuracy of line-of-sight (LOS) and NLOS under static conditions reaches 96.00%. Under different sampling steps, the average recognition accuracy of the three common obstacles, including wood boards, iron plates, and human body, is more than 95.63%. Compared with the origin, the average recognition accuracy of the proposed algorithm in this paper is improved by 18.87%.

ABSTRACT. Currently, there are numerous thorny issues in structured data and semi-structured full-text search scheme with large-scale text nature, like long data synchronization delay, inconvenient personalized business processing and low efficiency. To address these issues, this paper proposes an efficient algorithm based on Canal data synchronization framework and Elasticsearch full-text search engine. Firstly, we rewrite the Canal adapter component to obtain the flexible configuration of business data processing, thereby enhancing the secondary data processing ability of the framework and achieving the purpose of improving the efficiency of data synchronization. Secondly, by recording the synchronization time of nearby data in Canal framework, the weight of time series data is gradually decreased by combining with the exponential weighted average function to highlight the influence of recent data and present the novelty of data, which can achieve effective control the synchronization interval and duration by dynamically and flexibly setting the synchronization trigger period. Lastly, the Elasticsearch word tokenizer is modified, and then the configuration of custom expansion words and stop words dictionary are proposed to filter the query data effectively, thereby enhancing the query hit rate and accuracy. Extensive experiments on the data of traditional Chinese medicine demonstrate that the designed algorithm obtains high data synchronization efficiency, full text search speed and accuracy. Hence, the proposed algorithm is a milestone in smart healthcare.

ABSTRACT. Streaming feature selection (SFS), which enters the feature space in the form of flow and plays an important role in processing high dimensional data. Online streaming feature selection (OSFS) is a real-time streaming feature selection method. Most of the existing online streaming feature selection methods are based on the relevance between feature and label, ignoring the lack of label information, which frequently results in the failure of feature selection. In addition, in numerous practical application scenarios like intelligent medical treatment, label information with the high cost of acquiring label information is missing, which poses a great challenge to feature selection. To solve these problems, a novel unsupervised online streaming feature selection algorithm with density peak clustering (UOSFS-DPC) is proposed to reduce feature redundancy by clustering online streaming features. An online unsupervised relevance analysis and online unsupervised redundancy analysis are also adopted to select important feature subsets from the streaming features. Experimental results on six benchmark datasets and compared studies with eight representative SFS algorithms demonstrate that UOSFS-DPC outperforms its peers when label information is unknown.