Mixed Reality (MR) and Augmented Reality (AR) technologies have been used to improve remote collaboration. However, existing MR- or AR-based remote collaboration systems lack of a fully independent view sharing between the local user and remote user. This research propose a novel approach to enhance the remote collaboration using a drone and MR technology. By augmenting a virtual 3D avatar on the drone in the local environment, we propose the drone-driven agent to embody the remote user. And the view sharing between local and remote user is achieved by sending a real-time video stream of the local environment captured with the drone. There are three novelties including 1) fully independent view sharing, 2) augmenting virtual character on the drone to embody remote user, and 3) 3D AR sketching o facilitate communication between local and remote users. We implemented a proof-of-concept prototype to illustrate our design using a see-through type head-mounted display and a small-size drone. In addition, we provide discussion and implication for the future work to design drone-based remote collaboration systems.

— The rise of industrial progress has advanced the growth of deep learning (DL)-driven smart fault diagnosis techniques, particularly for condition-based maintenance (CBM). However, the training of these DL methods relies on large dataset, which is unrealistic to collect because fault signal is not practically viable in real case. To address this issue, this article proposes a conditional auxiliary classier cycle-consistent generative adversarial network restrained by Wasserstein distance with gradient penalty (CAC-CycleGAN-WGP). This model can generate superior-quality signals of the minority classes with stability from majority class. In the experimental section, a stacked autoencoder (AE)-based evaluator is proposed to evaluate the quality of these generated sample, and then imbalanced fault diagnosis is conducted at varying balance ratios based on two benchmarked datasets. The outcomes indicate that the proposed approach is adept at generating fault signals, leading to a notable enhancement in fault diagnosis accuracy as the generated samples are added. Additionally, the efficacy of the proposed framework was benchmarked against other commonly employed techniques. Among them, CAC-CycleGAN-WGP stands out with superior performance.

Augmented Reality (AR) has been used to enhance the shopping experience. However, existing AR shopping systems mainly focus on the solo user’s experience, while lacking multi-user experience. To address the gap, we propose a novel approach to collaborative shopping in a multi-user AR environment. By integrating the interactive personalized avatars of customers and shared gaze cues between shopping companions, we aim to understand how these technologies can enhance the collaborative shopping experience. We recruited thirty participants to conduct a 2 (personalized avatar: static vs. interactive) times 2 (shared gaze: without vs. with) within-subject repeat user study. The quantitative results from questionnaires showed that both interactive personalized avatars and shared gaze cues had positive effects on participants’ perceptions of enjoyment, usefulness, communication, co-presence, and future use. The combination of two features further enhanced the communication and perceived co-presence between shoppers and was preferred by participants. The qualitative results showed that interactive personalized avatars and shared gaze cues can enhance the shopping experience and promote efficiency which is consistent with quantitative results.

Throughout the development of Industry 5.0 towards a para-digm prioritizing human-centricity and sustainability, the potential of assembly lines with human-robot collaboration (HRC) is substantial. In HRC environments, human and robot operators share the workplace and can perform tasks simultaneously and collaboratively, amplifying work efficiency and operators’ welfare. This research investigates solutions to the assembly line balancing problem (ALBP) with HRC, where multiple human and robot operators work together. Using an adaptive simulated annealing (SA) framework for addressing ALBP with HRC, two innovative mechanisms are introduced-a new fitness value calculation method for roulette wheel selection and a pioneering heuristic approach. These mechanisms are devised to establish an innovative meta-heuristic approach based on SA for enhancing task allocation and resource management, improving productivity and operators’ well-being through strategic workload balancing between human and robot operators, and minimizing cycle times and the total number of operators required. The computational results using actual production data show that these mechanisms significantly enhance the solution quality, particularly in the large-size case study involving collaboration between multiple humans and robots in each workstation.

Vehicle routing problem (VRP) is one of the classic combinatorial optimization problems where an optimal tour to visit customers is required with a minimum total cost in the presence of some constraints. Recently, VRP is being solved with the use of deep reinforcement learning (DRL), with node sets considered (represented) as a graph structure. Existing Transformer based DRL solutions for VRP rely only on node information ignoring the role of information on the edges between nodes in the graph structure. In this paper, we proposed an attention-based end-to-end DRL model to solve VRP which embeds edge information between nodes for rich graph representation learning. We used Transformer based encoder-decoder framework with an edge information embedded multi-head attention (EEMHA) layer in the encoder. The EEMHA-based encoder learns underlying structure of the graph and generates an expressive graph topology representation by merging node and edge information. We trained our model using proximal policy optimization (PPO) with some code-level optimization techniques. We conducted an experiment on randomly generated instances and on a real-world data generated from road networks to verify the performance of our proposed model. The result from all experiments show that our model performs better than the existing DRL methods and most of the conventional heuristics with good generalizability from random instance training to real-world instance testing of different problem size. © 2023 Institute of Electrical Engineers of Japan. Published by Wiley Periodicals LLC.

Job shop scheduling problem (JSSP) is one of the well-known NP-hard combinatorial optimization problems (COPs) that aims to optimize the sequential assignment of finite machines to a set of jobs while adhering to specified problem constraints. Conventional solution approaches which include heuristic dispatching rules and evolutionary algorithms has been largely in use to solve JSSPs. Recently, the use of reinforcement learning (RL) has gained popularity for delivering better solution quality for JSSPs. In this research, we propose an end-to-end deep reinforcement learning (DRL) based scheduling model for solving the standard JSSP. Our DRL model uses attention-based encoder of Transformer network to embed the JSSP environment represented as a disjunctive graph. We introduced Gate mechanism to modulate the flow of learnt features by preventing noise features from propagating across the network to enrich the representations of nodes of the disjunctive graph. In addition, we designed a novel Gate-based graph pooling mechanism that preferentially constructs the graph embedding. A simple multi-layer perceptron (MLP) based action selection network is used for sequentially generating optimal schedules. The model is trained using proximal policy optimization (PPO) algorithm which is built on actor critic (AC) framework. Experimental results show that our model outperforms existing heuristics and state of the art DRL based baselines on generated instances and well-known public test benchmarks. © 2023 Institute of Electrical Engineers of Japan. Published by Wiley Periodicals LLC.

The rapid growth of electricity demands for recent years leads to many global concerns, including greenhouse, energy risk, and large size brownouts, which require modern energy management system to provide environmental-friendly power supply service with less cost and higher reliability. However, most of previous studies separately focus on improving prediction accuracy or reducing cost and emission of power supply solution by dispatch optimization. To exploit the benefits of microgrid system furthermore, this paper firstly proposes a comprehensive day-ahead multi-objective microgrid optimization framework that combines forecasting technology, demand side management (DSM) with economic and environmental dispatch (EED) together. Then, two versions of particle swarm optimization are implemented for obtaining load control plan from DSM model and power supply plan from EED model, respectively. Moreover, two different microgrids' applied scenarios are simulated with detailed sensitivities analysis on key parameters. Experiment results demonstrate effectiveness of the proposed framework, which can obtain load demands profile with better reliability, as well as power supply solution with less cost and lower emission. Meanwhile, beneficial decision supports are provided to manager for their references. © 2022 Institute of Electrical Engineers of Japan. Published by Wiley Periodicals LLC.

The flexible job-shop scheduling problem (FJSP) is a critical model in manufacturing systems that assigns operations from different jobs to various machines. However, optimizing multiple targets during the production scheduling process is always necessary. While the non-dominated sorting genetic algorithm (NSGA-II) is an effective method to solve the multi-objective FJSP, it can have the main drawbacks of converging too early and falling into local optimization. To address these issues, this research proposes a hybrid algorithm that combines NSGA-II and multi-objective simulated annealing using a pareto-domination based acceptance criterion (PDMOSA). The PDMOSA has a powerful search performance that can overcome the limitations of NSGA-II. The hybrid algorithm also includes original modification methods such as a deletion criterion, duplicated solution deletion, and new individual adding. Additionally, a plug-in decoding method is introduced. The proposed hybrid algorithm is compared with several improved ways based on NSGA-II in various experiments. The results demonstrate that the performance of the hybrid algorithm is better than the others in multi-objective FJSP.

The performance of a neural network relies on the depth of a model to learn the structural correlations of the features. Nevertheless, Graph Neural Networks (GNN) tends to lose its efficiency as the depth increases. In this paper we propose a technique to alleviate this problem by building on the existing GNN architecture. In effect, we installed a gating mechanism to overcome the propagation of noise information across the layers and trained the model using a proximal policy optimization (PPO), a policy gradient-based reinforcement learning algorithm. We trained the proposed model on a capacitated vehicle routing problem (CVRP) datasets generated on the fly. We used an encoder-decoder framework where the encoder learns the representation of the graph structured CVRP instance and the decoder learns to construct an optimal route based on the reward function designed. According to experiments using randomly generated test instances, the proposed model produces better results than the current deep reinforcement learning (DRL) methods to solve CVRP. To confirm the performance of our model, we also tested using locally generated real-world data parsed from digital maps. The results affirms that our model performs well in both random instance testing and real-world instance testing.

This study aims to enable jobs to go smoothly between shops on a production line by completing jobs in the upstream shop just in time (JIT) for the downstream shop. We propose solutions to a factory that is seeking ways for an upstream shop to complete every job at the precise time such that the downstream shop can process the job. We model the upstream shop as a flexible job shop and propose four methods that form a realtime scheduling and control system for JIT production. We first propose a method to set for each job a due date by which the job should be completed in the upstream shop. The due dates are set in such a manner that jobs would be completed JIT for the downstream shop, if they are completed JIT for their due dates. We then propose a method to estimate the minimum number of workers needed in the upstream shop for completing the jobs by their due dates. We further propose two methods that work dynamically to complete each job neither too early nor too late for its due date. One is a dispatching rule that dynamically sequences jobs in process according to urgency degree. The other is a job-selecting heuristic that dynamically assigns workers to jobs such that jobs not nearing completion will be given priority in processing. Simulations by using data from the factory show that the methods can achieve in real time (i.e. within 0.00 seconds) JIT production for a flexible job shop problem involving hundreds of operations. More extensive simulations by using a large number of randomly generated problem instances show that solutions obtained in real time by the proposed methods greatly outperform those obtained in much longer time by metaheuristics designed for solving similar problems, and that each proposed method outperforms its rivals in the literature. The findings imply that integrating fast and high-performing heuristics and rules can be a solution to solve large-scale scheduling problems in real time.

This paper is a study on an industrial engineering approach for the formulation of digital transformation (DX) strategies in the manufacturing industry. When a company is working on the formulation of a DX strategy, it is not uncommon for the stakeholders in the relevant departments to have different opinions and conflict with each other. Therefore, companies invite outside specialized consultants as facilitators and organize workshops with stakeholders to try to compile fair and objective DX requirements. However, not all of the stakeholders who participated in the workshop are satisfied with the results compiled by the consultant. In this paper, we propose to incorporate an augmented digital process of text mining for the content of stakeholders' requirements into the conventional real in-person discussion process of organizing requirements. To evaluate the usefulness of this proposal, we conducted an industrial experiment analyzing stakeholders’ opinions collected during a workshop at an industrial company. In other words, we evaluated whether or not it is possible to discover new insights and potential requirements that the consultant was not aware of. As a result of this experiment, we were able to confirm the possibility of revealing stakeholder opinions that were overlooked by the consultant during the workshop at the time, doing so by applying the proposed approach. From this, we concluded that incorporating text mining into the organization of stakeholder requirements would be effective in the requirements gathering process of DX strategy development.

The capacitated vehicle routing problem (CVRP), which is referred as NP-hard problem is a variant of Traveling Salesman Problem (TSP). CVRP constructs the route with the lowest cost without violating vehicle capacity constraints to meet demands of customer nodes. Following the advent of artificial intelligence and deep learning, the use of deep reinforcement learning (DRL) to solve CVRP is giving promising results. In this paper we proposed DRL model to solve CVRP. The transformer-based encoder of our proposed model fuses node and edge information to construct a rich graph embedding. The proposed architecture is trained using proximal policy optimization (PPO). Experiments using randomly generated test instances show that the proposed model gives rise to better results in comparison with the existing DRL methods. In addition, we also tested our model on locally generated real-world data to verify its performance. Accordingly, the results show that our model has a good generalization performance for both of random instance testing to real-world instance testing.

The rapid growth of electricity demand has led governments around the world to implement energy-conscious policies, such as time-of-use tariffs. The manufacturing sector can embrace these policies by implementing an innovative scheduling system to reduce its energy consumption. Therefore, this study addresses bi-objective job-shop scheduling with total weighted tardiness and electricity cost minimization under time-of-use tariffs. The problem can be decomposed into two sub-problems, operation sequencing and start time determination. To solve this problem, we propose a distributed-elite local search based on a genetic algorithm that uses local improvement strategies based on the distribution of elites. Specifically, chromosome encoding uses two lines of gene representation corresponding to the operation sequence and start time. We propose a decoding method to obtain a schedule that incorporates operation sequencing and start time. A perturbation scheme to reduce electricity costs was developed. Finally, a local search framework based on the distribution of elites is used to guide the selection of individuals and the determination of perturbation. Comprehensive numerical experiments using benchmark data from the literature demonstrate that the proposed method is more effective than NSGA-II, MOEA/D, and SPEA2. The results presented in this work may be useful for the manufacturing sector to adopt the time-of-use tariffs policy.

Multivariate time series forecasting has typically been a relevant and interesting topic in many fields, including economics, electricity consumption, solar energy, and traffic management. In these domains, owing to the complex dependencies among multiple variables and the mixed dependencies in the time dimension, it is challenging to forecast a multivariate time series precisely. Furthermore, most of the forecasting methods fail to capture the mixed influence of the different time-length dependencies among multiple variables. In this paper, a new deep learning framework is proposed for dealing with this challenging problem, named as mixed dependence time-series network (MDTNet). In this framework, stacked dilated convolutions and recurrent units are applied to extract the complex patterns in the long- and short-term mixed dependencies among multiple variables. The experiments show that our proposed framework yields significant results, outperforming the state-of-the-art baseline methods on three of the four benchmark datasets in large horizons and achieving a competitive performance in short horizons on all the benchmark datasets.

Sensor Data prediction has been an interesting and practical topic in many domains. In the process industry, sensor data prediction can help us detect, diagnose and even predict possible failures to reduce unnecessary losses. Due to the complex relationship among multiple sensors, it is challenging to accurately predict the time series of multivariate sensors. In this research, we aim to solve the problem of predicting the time series of several related sensor data and proposed a novel structure for addressing with this provocative problem. More specifically, several proposed mixed length dilation layers and recurrent cells are used to capture mixed length of time dependencies. Experiments demonstrate that our proposed model indicates competitiveness in predicting comparing with other baseline methods.

Most manufacturing industries are facing changes due to the increasing competition in the global market. Regarding this current situation, manufacturing firms that offer mass customization usually have a competitive edge over counterparts offering generic products. However, the Mass Production-to-Mass Customization (MP-to-MC) transition has brought upon unprecedented challenges to many Small and Medium Enterprises (SMEs). One of the challenges is to identify customized products in harsh production environments. The reason behind this is that identification tags such as barcodes, Quick Response (QR) codes and Radio Frequency Identification (RFID) cannot function in special production processes like heating and dissolution. It is therefore of prime importance to find a solution by devising a product identification method without making use of marks or tags. In the paper, a novel method that using computer vision to identify the customized products in mass customization and a hybrid Convolutional Neural Network (CNN) model are proposed. To illustrate the efficacy of the proposed method, a case study in a shoe-manufacturing company was reported. The results yielded demonstrated that the proposed method is an efficient and economical solution.

Increasing energy demand can create undesired problems for many governments worldwide. Several policies, such as time-of-use (TOU) tariffs, have been put in place to overcome such demand. The TOU policy's objective is to reduce electrical load during peak periods by shifting the use to off-peak periods. To that end, this paper addresses the bi-objective permutation flow-shop scheduling, minimizing total weighted tardiness and electricity costs. We propose a meta-heuristic algorithm based on SPEA2 to solve the problem. We conducted numerical experiments to evaluate the efficacy of the proposed algorithm by comparing it with NSGA-II. The results show that the proposed approach was more efficient compare with NSGA-II.

The genetic algorithm (GA) is one of the most common solutions to solve many optimisation problems. Its distributed version, Island Model GA (IMGA), was introduced to overcome more complex and scalable cases. However, there is a recurrent problem in IMGA called premature convergence as a consequence of selection in the migration. This process is a mechanism of migrating individuals from one into another island to keep population diversity. The primary cause is the structural similarity of a migrated individual because of the genetic operator configurations are identical. Localised IMGA (LIMGA) tries to implement different island characteristics to avoid premature convergence. The main motivation of this paper is to investigate the performance of LIMGA capability in maintaining population diversity. In detail, the contributions of this research are (1) to prove LIMGA concept in handling general optimisation problem, (2) to analyse the performance LIMGA in diversity preservation, and (3) compare LIMGA performance with the current solvers. By harmonising three different GA cores, LIMGA could overcome computationally expensive functions with a great result and acceptable execution time. Moreover, because of its success in maintaining the diversity, Localised Island Model Genetic Algorithm (LIMGA) could lead to the among other current solvers for this case.

This study is aimed to smooth the job flow between consecutive production shops along a production line. We propose solutions to a factory that wants an upstream shop to complete every job at a timing that the downstream shop can process the job. We model the upstream shop as a reentrant flexible job shop and propose three online executable methods forming a control system. One is a due date setting method that estimates the flow time of a job by making a quick schedule of jobs in bottleneck workstations. Another is a dispatching rule that sequences jobs according to urgency degree. The other is a heuristic that controls the start time of processing the final operation in a job. Simulations by using data from the factory show that each method is superior to its rivals in the literature, and when working together, the methods can not only complete jobs at the desired timings but also reach instantly far better solutions than metaheuristics customized for solving similar problems.

Electricity consumption forecasting plays an important role in establishing and maintaining electric supply management systems. Power companies need to keep a balance between the power demand and supply for customers; this requires an accurate forecast. However, electricity consumption forecasting is affected by various factors such as different weather conditions, season, or temperature. If we cannot predict electricity accurately, the balance between the demand and supply would be destroyed, which may cause huge penalties to power companies. Therefore, electricity consumption forecasting is an important task. The purpose of this study was to forecast the electricity consumption of a manufacturing company every half an hour in the next day to prevent a power supply company from running out of power. In our work, we proposed a short-term electricity consumption forecasting method based on the attentive encoder-decoder and several nonlinear multi-layer correctors. The proposed method is verified in several experiments by using the actual data on electricity consumption of the manufacturing company. The results show that the proposed method outperforms previous methods.

The University Course Timetabling Problem (UCTP) is a scheduling problem of assigning a teaching event in a certain time and room by considering the constraints of university stakeholders such as students, lecturers, and departments. This problem becomes complicated for universities with a large number of students and lecturers. Moreover, several universities are implementing student sectioning, which is a problem of assigning students to classes of a subject while respecting individual student requests, along with additional constraints. Such implementation also implies the complexity of constraints, which is larger accordingly. However, current and generic solvers have failed to meet the scalability and reliability requirements for student sectioning UCTP. In this paper, we introduce the localized island model genetic algorithm with dual dynamic migration policy (DM-LIMGA) to solve student sectioning UCTP. Our research shows that DM-LIMGA can produce a feasible timetable for the student sectioning problem and get better results than previous works and the current UCTP solver. Our proposed solution also consistently yield lower violation number than other algorithms, as evidenced by UCTP benchmark experiment results. © 2019 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.

Energy demand is increasing as the population and economy grow. Many countries have implemented time-of-use (TOU) tariffs to meet such demand so that the demand during peak periods could be reduced by shifting its usage from peak periods to off-peak periods. This paper addresses the unrelated parallel machine scheduling under TOU to minimize the sum of weighted makespan and electricity cost. Because the problem has nonregular performance measure, delaying the starting time of the job can produce a better solution. Hence, not only do we determine the job sequencing and the job assignment, but also we determine the starting time of the job. We propose a triple-chromosome genetic algorithm that represents the job sequencing, the job assignment and the optimal starting time of the job simultaneously. A self-adaptive algorithm is developed to determine the value of the third chromosome after crossover and mutation process. Numerical experiment and statistical analysis are conducted to show the appropriateness and efficacy of the proposed approach. © 2019 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.

Island Model Genetic Algorithm (IMGA) is a multi-population based GA. IMGA aimed to avoid local optimum by maintaining population (island) diversity using migration. There are several mechanisms of migration and individual selection such as the best (or worst) individual selection, new naturally inspired evolution model, and dynamic migration policy. Migration can delay island (local) convergence and intrinsically preserve diversity. Ironically, migration is also potential to bring overall island (global) convergence, faster. In a certain generation, the migrated individuals among islands will have similar value (genetic drift). So, this work aims to preserve global diversity better by implementing Localized IMGA (LIMGA) and Dual Dynamic Migration Policy (DDMP). LIMGA creates unique evolution trends by using a different kind of GAs for each island. DDMP is a new migration policy which rules the individual migration. DDMP determines the state of an island according to its diversity and attractivity level. By determining its states, DDMP ensures the individual migrating to the correct island dynamically. We call the combination of LIMGA and DDMP as Dual Migration LIMGA (DM-LIMGA). Our experiments show that DM-LIMGA can preserve the diversity better. As its implication, DM-LIMGA can create a more extensive search space and dominates the results among other solvers.

Due to new government legislation, customers’ environmental concerns and continuously rising cost of energy, energy efficiency is becoming an essential parameter of industrial manufacturing processes in recent years. Most efforts considering energy issues in scheduling problems have focused on static scheduling. But in fact, scheduling problems are dynamic in the real world with uncertain new arrival jobs after the execution time. This paper proposes an energy efficient dynamic flexible flow shop scheduling model using the peak power value with consideration of new arrival jobs. As the problem is strongly NP-hard, a priority based hybrid parallel Genetic Algorithm with a predictive reactive complete rescheduling strategy is developed. In order to achieve a speedup to meet the short response in the dynamic environment, the proposed method is designed to be highly consistent with the NVIDIA CUDA software model. Finally, numerical experiments are conducted and show that our approach can not only solve the problem flexibly, but also gain competitive results and reduce time requirements dramatically.

This paper proposes a framework of short term and intensive workshop facilitation for multi-party stakeholders in PLM strategy planning phase. We have been empirically pursuing what a valuable facilitation for the workshop is; how multi-party PLM stakeholders can build proactively a mutual consensus in as short a time as possible. PLM project promotion members always encounter a difficulty of consensus building. This is because various stockholders have different opinions and responsibilities through sales, engineering, manufacturing, and service departments. Firstly, we mention key challenges of multi-party consensus building in PLM strategy planning phase. Secondly, we propose a programmatic framework on intensive workshop-facilitation which is configured twelve steps. The key outcome of the workshop is to craft a PLM Success Value Roadmap (PSVR) which is contained various hypothesis defined by the workshop participants helping by facilitators (KPIs). For example, there are PLM vision, strategy, initiative, process, and key performance indicator. Thirdly, we mention an empirical case study conducted our proposed workshop-facilitation method for an industrial company. Seventeen stakeholders were joined as the workshop participants who were invited from three different business units. It was held as a two-day intensive PLM trial workshop. Finally, we found that the proposed workshop-facilitation as a consensus building method contributed to the satisfaction of more than 60% of the participants. 85% of the participants commented that they would encourage colleagues to participate in the workshop that we have developed. We conclude that the multi-party intensive workshop was a valuable experience that it allows stakeholders to produce a PLM strategy in a relatively short time.

The University Course Timetabling Problem (UCTP) is a scheduling problem of assigning teaching event in certain time and room by considering the constraints of university stakeholders such as students, lecturers, departments, etc. This problem becomes complicated for universities which have immense number of students and lecturers. Therefore, a scalable and reliable timetabling solver is needed. However, current solvers and generic solution failed to meet several specific UCTP. Moreover, some universities implement student sectioning problem with individual student specific constraints. This research introduces the Reinforced Asynchronous Island Model Genetic Algorithm (RIMGA) to optimize the resource usage of the computer. RIMGA will configure the slave that has completed its process to helping other machines that have yet to complete theirs. This research shows that RIMGA not only improves time performance in the computational execution process, it also offers greater opportunity to escape the local optimum trap than previous model.

Nowadays, energy saving is one of the most talked about issues in our life, it is also increasingly important in the manufacturing industry. This research considers the dynamic flexible flow shop scheduling (DFFS) problem, which is an extended version of the classical flow-shop scheduling problem. A flexible flow shop has multiple stages with multiple machines at each stage for processing multiple products. Previous research on DFFS aimed to achieve just-in-time production, or reducing difference between the actual completion time and the due date of each job. However, little research has been made on energy saving of machines in production. To address such a need, this paper proposes a method that dynamically turns on and off machines so as to reduce energy consumption while achieving JIT production. The proposed method has been tested on different environments, and the results show that it is high performing for both JIT production and energy saving.

With the rapid growth of electricity demand, many governments around the world have implemented the energy-conscious policy such as time-of-use policy. This paper addresses a bi-objective single machine scheduling with the total tardiness and electricity cost minimization under time-of-use tariffs. The problem is formulated as a mixed integer programming model. The CPLEX solver solves a small size instance to validate the model. We also describe the procedure to obtain the set of non-dominated solution using commercial solver. The complexity of the model is tested on several problem instances. The results show that the problem is hard to solve even for medium size instances. Hence, we propose a genetic algorithm with random insertion to obtain the set of Pareto solutions for large instances.

Electricity consumption forecasting has a main role in the energy supply management system of a power supply company. A power supply company needs to keep the balancing of the electricity demand and supply for their customers. The target is to forecast the electricity consumption in manufacturing company for each 30-minutes in the next day to prevent the lack of electricity supply from a power supply company. Due to this problem, it is the challenge for short term electricity time series consumption forecasting. In this work, we proposed the model of deep learning neural network with approach the combination of Long Short-Term Memory (LSTM) and Feed Forward Neural Network (FFNN) to perform the electricity forecasting. This proposed method (LSTM-FFNN) was implemented in the time-series data of electricity consumption on a manufacturing company. In our experiment, we used LSTM to perform the time-series forecasting by using historical data of electricity consumption, and we performed FFNN along with additional information which represented by one-hot encoding shape to increase the forecasting performance. Experimental results showed that LSTM-FFNN gave the better result as we compared with our baseline which is the original LSTM and Moving Average (MA) based on the Root Mean Squared Error (RMSE) score.

On January 11, 2017, Waseda University launched a university–industry collaboration program with a local government body in Kitakyushu City, Japan and 20 local businesses in the area. Officially called the “Waseda University IPS Kitakyushu Consortium (IPSKC),” the program aims to change the direction of local society and to develop innovative business and technology solutions in the era of Industry 4.0. In addition, it seeks to contribute to future global business development with neighboring Asian countries. As a first step, one of the program’s key initiatives was to offer consortium members an Internet of Things (IoT) business modeling workshop as part of the University’s industrial IoT/business engineering education program. This paper discusses an approach to facilitate workstream as a case study. We present the key results of the workshop; and discuss the future outlook of IoT and PLM education program.

Years after being firstly introduced by Fraser and remodeled for modern application by Bremermann, genetic algorithm (GA) has a significant progression to solve many kinds of optimization problems. GA also thrives into many variations of models and approaches. Multi-population or island model GA (IMGA) is one of the commonly used GA models. IMGA is a multi-population GA model objected to getting a better result (aimed to get global optimum) by intrinsically preserve its diversity. Localization strategy of IMGA is a new approach which sees an island as a single living environment for its individuals. An island’s characteristic must be different compared to other islands. Operator parameter configuration or even its core engine (algorithm) represents the nature of an island. These differences will incline into different evolution tracks which can be its speed or pattern. Localization strategy for IMGA uses three kinds of single GA core: standard GA, pseudo GA, and informed GA. Localization strategy implements migration protocol and the bias value to control the movement. The experiment results showed that localization strategy for IMGA succeeds to solve 3-SAT with an excellent performance. This brand new approach is also proven to have a high consistency and durability.

Genetic algorithm (GA) is one of the standard solutions to solve many optimization problems. One of a GA type used for solving a case is island model GA (IMGA). Localization strategy is a brand-new feature for IMGA to better preserves its diversity. In the previous research, localization strategy could carry out 3SAT problem almost perfectly. In this study, the proposed feature is aimed to solve real parameter single objective computationally expensive optimization problems. Differ with an issue in previous research which has a prior knowledge and binary, the computationally expensive optimization has not any prior knowledge and floating type problem. Therefore, the localization strategy and its GA cores must adapt. The primary goal of this research is to analyze further the localization strategy for IMGA's performance. The experiments show that the new feature is successfully modified to meet the new requirement. Localization strategy for IMGA can solve all computationally expensive functions consistently. Moreover, this new feature could make IMGA reaches leading ratio 0.47 among other current solvers.

Genetic algorithm (GA) is one of the standard solutions to solve many optimization problems. One of a GA type used for solving a case is island model GA (IMGA). Localization strategy is a brand-new feature for IMGA to better preserves its diversity. In the previous research, localization strategy could carry out 3SAT problem almost perfectly. In this study, the proposed feature is aimed to solve real parameter single objective computationally expensive optimization problems. Differ with an issue in previous research which has a prior knowledge and binary, the computationally expensive optimization has not any prior knowledge and floating type problem. Therefore, the localization strategy and its GA cores must adapt. The primary goal of this research is to analyze further the localization strategy for IMGA's performance. The experiments show that the new feature is successfully modified to meet the new requirement. Localization strategy for IMGA can solve all computationally expensive functions consistently. Moreover, this new feature could make IMGA reaches leading ratio 0.47 among other current solvers.

Today, many of manufacturing system integrators are pursuing new business solution on information technology (IT) embedding Internet of Things (IoT) functionalities. However, how about conventional type of system integrators? Although their customers already require new innovative IoT solutions, such integrators tend to still provide traditional operational technology (OT) solutions that the customers are less interested in. This is a cause of lack of knowledge of IT technology that OT engineers have never experienced. This paper introduces a preliminary study for a method development on pragmatic workshop combining with business model definition and IoT technology training for such OT system integrators. This workshop method particularly focuses on convergence of OT and IT; defining a new business value chain and experiencing of commercial IoT technology as hands-on session. This paper is mainly discussed a real case study as a preliminary trial in an industrial city in Far East region.

Unrelated parallel machine scheduling under time-of-use electricity price is addressed in this paper. In this setting, price of electricity can be different among various periods of the day. The objective is to minimize total cost consisting of makespan cost and electricity cost. Genetic algorithm (GA) is used to solve the unrelated parallel machine scheduling under time varying tariffs. Chromosome decoding, inspired by greedy total cost, is proposed to transform individual into feasible schedule. Furthermore, generated schedule from the individual is improved by job delay mechanism that shifts jobs to other periods to avoid high electricity cost. Finally, numerical experiment is conducted to implement the approach. Preliminary result shows that our proposed approach is effective and efficient to solve the corresponding problem.

The common problem in island model is the way to migrate individual from one to another island, or usually called as migration policy. Previous researches in this movement protocol could be categorized into two different approaches, diversity preservation based such as migration protocol in LIMGA and better island pursuing based such as new dynamic migration policy. The main purpose of this works is to introduce a brand-new migration mechanism called as Dual Dynamic Migration Policy (DDMP) for island model GA. DDMP will take the advantage from result pursuer of dynamic migration policy and convergence avoider of LIMGA's migration protocol. The experiment result shows that DDMP could give great result while carrying out the general optimization cases. It could produce the best score result for all cases among previous island model migration methods. This work also compares DDMP with the-best-known-so-far solution for the problem set.

The global manufacturing business is in the new era of industrial revolution based on digital data across the whole business processes. Internet of Things (IoT) is one of extremely high expected technologies. They contribute product lifecycle management (PLM) process, such as remote monitoring of field service and predictive quality reliability engineering design. However, it assumes significant difficulties for small and medium sized enterprises (SME) to launch rapidly IoT solution for their business efficiency or strategic differentiation. Thus, this paper proposes a pragmatic IoT Innovation workshop approach for such SMEs’ employees. This is as an industry-university collaboration PLM educational program utilizing both design thinking business methodology and commercial IoT application technology hands-on. It also introduces outcomes as a preliminary phase for empirical study on this workshop approach that deployed for a local city in the Far East area.

Nowadays, energy conservation is one of the most talked about issues in our life, it is also increasingly important in dynamic flexible flow shop scheduling problems. Previous research did much on how to minimize total earliness and tardiness of all jobs, or to achieve just-in-time production, whereas little effort has been made on how to save energy. Therefore, to address such need, a dynamic machine switching method has been proposed in this paper. The main purpose is to minimize both energy consumption and just-in-time production objectives. Featured by switching machines dynamically, the proposed method has been tested on different environments, whose results show that it has good performance for energy conservation.

Job shop scheduling problem (JSP) is a strongly NP-hard combinatorial optimization problem. It is difficult to solve the problem to the optimum in a reasonable time. Teaching-learning-based optimization (TLBO) algorithm is a novel population oriented meta-heuristic algorithm. It has been proved that TLBO has a considerable potential when compared to the best-known heuristic algorithms for scheduling problems. In this paper, the traditional TLBO is improved to enhance diversification and intensification when exploring solutions for JSP. The improvements include changing the coding method, increasing number of teachers, introducing new learners and performing local search around potentially optimal solutions. To show effectiveness of the improved TLBO algorithm, the simulation results obtained by the improved TLBO for benchmark problems are compared with results obtained by the traditional TLBO and the best known lower bounds.

Competition and resultant complexity of today's production industry require that enterprises realize the importance of reducing total costs in production. Uncertainties of customer orders also make it difficult to determine the schedule for minimizing inventories, earliness and tardiness. In this paper, we propose a new production model for flexible flow shops, which aims to reduce the total inventories, earliness and tardiness. The model divides a flexible flow shop (FFS) into a make-to-stock (MTS) part and a make-to-order (MTO) part by applying a decoupling point. In the MTS part, jobs are manufactured into semi-finished products and then stored as inventories at the decoupling point. As soon as a customer order is received, the inventories are released into the system, starting undergoing processing in the MTO part. This shortens the lead time for manufacturing the products. Another advantage of the proposed model is that the less number of operations in the MTO part than that in a job makes it easier to avoid tardiness. In addition, we designed two types of models: dynamic and static, which depends on whether the decoupling point is dynamically adjusted to adapt to different arrival rates of customer orders. The reason why we design two types is to compare the performance of the proposed two models. Results show that the dynamic hybrid model outperforms pull, push and static hybrid models for reducing costs.

Flexible job-shop scheduling problem (FJSP) is an extended job-shop scheduling problem. FJSP allows an operation to be processed by several different machines. FJSP with overlapping in operations means that each operation is divided into several sublots. Sublots are processed and transferred separately without waiting for the entire operation to be processed. In previous research, a mathematical model was developed and a genetic algorithm proposed to solve this problem. In this study, we try to improve the procedure of previous research to achieve better results. The proposed improvements were tested on some benchmark problems and compared with the results obtained by previous research.

In this paper, we aim to solve the resource-constrained multiproject scheduling problem (rc-mPSP), in which more than one project are scheduled simultaneously, projects share global resources, and the average project delay and total project time are minimized as objectives. In order to solve this problem by a centralized scheduling method, we present a new genetic algorithm (GA) approach. In this procedure, we follow the GA described in Okada et al.. (2014) and improve its genetic operators, such as crossover and mutation, and local search so as to work better on rc-mPSP. Furthermore, in order to evaluate the proposed approach, we compare the results of the computational experiment on certain standard project instances with the several competing decentralized methods and centralized methods presented in the literature. We show that our procedure is one of the most competitive among such algorithms. (c) 2014 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.

In this paper, we aim to solve the resource-constrained multiproject scheduling problem (rc-mPSP), in which more than one project are scheduled simultaneously, projects share global resources, and the average project delay and total project time are minimized as objectives. In order to solve this problem by a centralized scheduling method, we present a new genetic algorithm (GA) approach. In this procedure, we follow the GA described in Okada et al.. (2014) and improve its genetic operators, such as crossover and mutation, and local search so as to work better on rc-mPSP. Furthermore, in order to evaluate the proposed approach, we compare the results of the computational experiment on certain standard project instances with the several competing decentralized methods and centralized methods presented in the literature. We show that our procedure is one of the most competitive among such algorithms. (c) 2014 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.

This study intends to help manufacturers that use flexible job shops improve performance of due date assignment, that is, setting delivery times to jobs that arrive dynamically. High performing due date assignment enables achieving on-time delivery and quick response of delivery time to customer orders. Traditional methods for due date assignment are predefined equations that estimate the duration of making a product in the production system. Such equations are sufficient for relatively simple systems such as single machine shops, but are not very high in accuracy for complex systems such as flexible job shops. To improve due date assignment for such systems, we propose a more flexible method that uses distributed learning to learn the remaining time of a job inside the system. We let each workstation in the production shop be a distributed unit that updates its local queuing time and interacts with other units to provide the total remaining time of a job. We carry out extensive computational experiments to evaluate performance of the proposed method, and the results show that it outperforms two advanced equational methods in terms of both accuracy of estimation and stability in performance.

© 2015, IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved.To reduce inventory costs and increase customer satisfaction in a real production environment, Just-in-Time (JIT) manufacturing should be combined with more flexibility. Flexibility in manufacturing system ensures improved product quality and mass customization. In this paper, we consider flexible job-shop problem (FJSP) to include several types of flexibility. Prior research on JIT scheduling problems paid little attention to flexible job shop systems. In this paper, an optimization method for FJSP for JIT manufacturing was proposed, genetic algorithm and time control are integrated to improve the JIT performance. Computational simulations and comparisons demonstrate that the proposed method shows more competitive performance than other evolutionary algorithms.

We consider the problem of estimating lead times of orders that arrive dynamically in a flexible flow shop. When an order arrives, we estimate the time between its arrival and its completion. Good estimation would help improve on-time delivery. Existing estimation methods are mostly designed based on an order's processing time and the shop's inventory status. We improve the accuracy of estimation by developing new methods that further take into account the characteristics of the shop system such as parallelization of queues. We also consider the bottleneck effect, which might largely dominate the waiting time in the system. Results of computational experiments show that the proposed methods outperform existing methods in terms of some accuracy measures.

We consider the problem of estimating flow times of jobs that arrive dynamically in a manufacturing system. A job's flow time refers to the time between the job's arrival and completion. Most existing methods use some predefined equations for such estimation, and most of the equations are designed for single machine manufacturing systems. To better estimate the flow time of a job in a more complex system in which there are multiple machines and multiple workstations, we propose a distributed learning approach that divides the manufacturing system into multiple small parts and collects real-time local information in each part to predict the waiting time for a job. We evaluate the proposed approach by comparing it with existing methods using a variety of problem instances. The results show that the proposed approach outperforms existing methods and accordingly might improve the level of customer service when being used for due date promising.

This article proposes a novel method to solve the manufacturing scheduling problems in multi-agent system (MAS) for dynamic environment. The study focuses on the machine selection work for jobs and proposes a remaining time prediction method, which will help a job choose machines based on estimated finishing time. In this way, we can keep the job finishing just-in-time, which aims to support a customer-oriented lead time policy, and can provide better manufacturing performance and increase customer service level. This article provides experimental results compared with another approach in previous research, through which we can find that in some case if we put the Just-in-time philosophy into consideration, the proposed method will deliver competitive performance.

Currently, expectations from Internet of Things (IoT) are extremely high worldwide. However, mechanical engineers working on product design processes do not always benefit. This paper utilizes the up-to-date IoT-enabled commercial product lifecycle management (PLM) system to examine the effectiveness of IoT value for mechanical engineers. Moreover, it discusses the maturity model for digital engineering for the new era of WT. Furthermore, this paper introduces significant initiatives such as a PLM functionality connected to everything, monitoring of field data for getting back to design process, and virtual 3D data and real field data superposition. This paper concludes that there is high potential to use IoT technology for mechanical engineers from the perspective of quality and reliability design innovations.

© 2015, IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved. To reduce inventory costs and increase customer satisfaction in a real production environment, Just-in-Time (JIT) manufacturing should be combined with more flexibility. Flexibility in manufacturing system ensures improved product quality and mass customization. In this paper, we consider flexible job-shop problem (FJSP) to include several types of flexibility. Prior research on JIT scheduling problems paid little attention to flexible job shop systems. In this paper, an optimization method for FJSP for JIT manufacturing was proposed, genetic algorithm and time control are integrated to improve the JIT performance. Computational simulations and comparisons demonstrate that the proposed method shows more competitive performance than other evolutionary algorithms.

To reduce inventory costs and increase customer satisfaction in a real production environment, Just-in-Time (JIT) manufacturing should be combined with noire flexibility. Flexibility in manufacturing system ensures improved product quality and mass customization. In this paper, we consider flexible job-shop problem (FJSP) to include several types of flexibility. Prior research on JIT scheduling problems paid little attention to flexible job shop systems. En this paper, an optimization method for FJSP for JIT manufacturing was proposed, genetic algorithm and tinge control are integrated to improve the JIT performance. Computational simulations and comparisons demonstrate that the proposed method shows more competitive performance than other evolutionary algorithms. (C) 2015, IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved.

This study proposes an efficient mixed integer linear programming (MILP) model for a generalized job shop production process. In the process, a workstation contains one or multiple machines and each job visits some of the workstations in a specific sequence. It is allowed that a job visits the same workstation more than once. Although some similar processes were modeled by MILP in previous studies, the models are unable to solve problems that involve more than ten jobs due to high computational complexity. Our proposed model outperforms the best model that is identified among 23 research papers with regard to computational complexity. Simulation results show that our model is able to solve a problem with a dozen of jobs, which is classified as a large-scale problem in the literature.

Although many production scheduling systems are developed, it is difficult to introduce for medium-sized enterprises, because of high cost of initial and maintenance investment. However functions provided by it, such as available-to-promise under the restrictions of the own process, resource leveling and so on, are required by all of enterprises to make their profits. To realize a scheduling system, which can be introduced more easily for the small and medium-sized enterprises, a novel spiral-evolutional scheduling system is proposed in this paper. This system is getting started with gathering of production information in factories, and through progress monitoring a scheduling function is installed and evolved in a spiral evolutional systemizing process. In this paper, the concept of the spiral-evolutional production scheduling system is described, and the implementation of this system and the merit of introduction of it are shown.

In this paper, we aim to solve the problem of resource-constrained project scheduling with multiple modes (rc-PSP/mM), in which multiple execution modes are available for each of the project's activity and with minimization of makespan as objective. We present a new genetic algorithm approach in order to solve this problem. In this procedure, we propose a new mutation operator that exploits a critical path and two new local search procedures, i.e. critical path improvement local search (cpiLS) and iterative forward/backward local search (ifbLS), using activity list characteristics. The cpiLS reduces the critical path and the ifbLS improves resource allocation of the schedule of rc-PSP/mM. Also, to evaluate the proposed approach, we compare the results of the computational experiment on certain standard project instances with the several competing heuristic procedures presented in the literature, and it has been revealed that our procedure is one of the most competitive among the algorithms. (c) 2014 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.

The semiconductor final testing scheduling problem (SFTSP) is a variation of the complex scheduling problem, which deals with the arrangement of the job sequence for the final testing process. In this paper, we present an actual SFTSP case includes almost all the flow-shop factors as reentry characteristic, serial and batch processing stages, lot-clusters and parallel machines. Since the critical equipment needs to be utilized efficiently at a specific testing stage, the scheduling arrangement is then playing an important role in order to reduce both the makespan and penalty cost of all late products in total final testing progress. On account of the difficulty and long time it takes to solve this problem, we propose a multi-objective optimization approach, which uses a lot-merging procedure, a new job-based encoding method, and an adjustment to the non-dominated sorting genetic algorithm II (NSGA-II). Simulation results of the adjusted NSGA-II on this SFTSP problem are compared with its traditional algorithm and much better performance of the adjusted one is observed. © (2013) Trans Tech Publications, Switzerland.

In semiconductor manufacturing factories, the process of wafer fabrication is the most technologically complex and capital intensive stage. This process is configured as a reentrant flow shop process with many machines and processing steps. It needs an efficient and effective scheduling method for large size process in order to increase the competitiveness. The reentrant flow shop problem (RFSP) means that all jobs have the same route through the shop machines and the same shop machine is used several times to complete a job. This research provides an effective fuzzy-based multi-term genetic algorithm to solving RFSP with the objective of minimizing the total turn around time (TTAT). The proposed method focuses on the critical point in scheduled solutions. The middle position of longest TAT is defined as the critical point. According to the critical point and current generation, fuzzy logic chooses the focused term of chromosome, then the genetic algorithm effects on this term. In each evolution, only corresponded part of chromosome is evolved by crossover and mutation while other parts of chromosome remain unchanged. Through computational experiments, the effectiveness of the fuzzy-based multi-term genetic algorithm is evaluated.

In production processes, just-in-time (JIT) completion of jobs helps reduce both the inventory and late delivery of finished products. Previous research which aims to achieve JIT job completion mainly worked on static scheduling problems, in which all jobs are available from time zero or the available time of each job is known beforehand. In contrast, dynamic scheduling problems which involve continual arrival of new jobs are not much researched and dispatching rules remain the most frequently used method for such problems. However, dispatching rules are not high-performing for the JIT objective. This study proposes several routing strategies which can help dispatching rules realize JIT completion for jobs arriving dynamically in hybrid flow shops. The routing strategies are based on distributed computing which makes realtime forecast of completion times of unfinished jobs. The advantages include short computing time, quick response and robustness against disturbance. Computer simulations show that the performance of dispatching rules combined with the proposed routing strategies is significantly higher than that of dispatching rules only and that of dispatching rules combined with the previous routing methods. (c) 2012 Elsevier Ltd. All rights reserved.

In production processes, just-in-time (JIT) completion of jobs helps reduce both the inventory and late delivery of finished products. Previous research which aims to achieve JIT job completion mainly worked on static scheduling problems, in which all jobs are available from time zero or the available time of each job is known beforehand. In contrast, dynamic scheduling problems which involve continual arrival of new jobs are not much researched and dispatching rules remain the most frequently used method for such problems. However, dispatching rules are not high-performing for the JIT objective. This study proposes several routing strategies which can help dispatching rules realize JIT completion for jobs arriving dynamically in hybrid flow shops. The routing strategies are based on distributed computing which makes realtime forecast of completion times of unfinished jobs. The advantages include short computing time, quick response and robustness against disturbance. Computer simulations show that the performance of dispatching rules combined with the proposed routing strategies is significantly higher than that of dispatching rules only and that of dispatching rules combined with the previous routing methods. (c) 2012 Elsevier Ltd. All rights reserved.

For manufacturers, the integration of high performance manufacturing with customer-oriented practices plays an important role in improving the performance of their business system. The benefits from such integration can only be maximized when the two parts are designed to work cooperatively. Though previous research has contributed much to manufacturing control algorithms and customer service practices, there has been little consideration of the two parts as a whole; consequently, the methods proposed may not be well supported by the other practices adopted in the system. This study develops production control methods that support a customer-oriented lead time policy, and aims to increase the performance of both manufacturing and customer service. The control methods are proposed for hybrid flow shops handling orders arriving dynamically. Computer simulations are conducted on a large number of problem instances, and the results show that the designed distributed feedback and decision-making functions enable the proposed methods to significantly outperform existing methods in achieving just-in-time (JIT) job completion under customized product lead times. Even taking into account the possible tradeoff between JIT job completion and flow time length, the proposed methods still deliver competitive performance. (C) 2011 Elsevier B.V. All rights reserved.

Based on the concept and principles of quantum computing, this paper proposes a new evolutionary algorithm called multiupdate mode quantum evolutionary algorithm (MMQEA). MMQEA, like the other classic evolutionary algorithms, is also characterized by the representation of the individual evaluation function and the population dynamics; however, instead of binary, numeric, or symbolic representation, MMQEA uses two interactional q-bit strings as an individual. Update modes are introduced as a variational operation that evolves the individuals to make a better solution. The proposed individual structure and update modes are inspired by quantum entanglement. Update modes perform as reproducing the states of a pair of q-bit strings of individual simultaneously. For guiding the individual evolution to maintain the population diversity and avoid prematurity, each q-bit string of individual provides its evolutionary history information to another. To demonstrate its effectiveness and applicability, the proposed algorithm was tested on two famous combinatorial optimization problems, namely, the knapsack problem and flow shop problem. The results show that MMQEA performs very well compared to quantum evolutionary algorithm (QEA) and the conventional genetic algorithm. (C) 2012 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.

This paper proposed a multi-objective local search procedure (MOLS). It is combined with NSGA-II for solving bi-criteria PFSP with the objectives of minimizing makespan and maximum tardiness. Utilizing the properties of active blocks for flow shop scheduling problem, neighborhood structures MOINS (multi-objective insertion) and MOEXC (multi-objective exchange) are designed in order to improve efficiency of perturbation. Any perturbation based on MOINS and MOEXC takes effect on different criteria simultaneously. The original idea of MOLS is systematic change neighborhoods in the local search procedure. The search direction of MOLS on an individual is naturally guided by interaction of MOINS and MOEXC. Moreover, there is no need to set parameters in MOLS. The MOLS combined with popular multi-objective evolutionary algorithm NSGA-II (Non-dominated Sorting Genetic Algorithm-II) is called as "NSGA-II- MOLS". To illustrate the efficacy of proposed approach, four different scaled problems are used to test performance of NSGA-II-MOLS. The numerous comparisons show efficacy of NSGA-II-MOLS is better than most of algorithms even with the same number of individual evaluations and parameters setting. © 2012 The Institute of Electrical Engineers of Japan.

Based on the concept and principles of quantum computing, this paper proposes a new evolutionary algorithm called multiupdate mode quantum evolutionary algorithm (MMQEA). MMQEA, like the other classic evolutionary algorithms, is also characterized by the representation of the individual evaluation function and the population dynamics
however, instead of binary, numeric, or symbolic representation, MMQEA uses two interactional q-bit strings as an individual. Update modes are introduced as a variational operation that evolves the individuals to make a better solution. The proposed individual structure and update modes are inspired by quantum entanglement. Update modes perform as reproducing the states of a pair of q-bit strings of individual simultaneously. For guiding the individual evolution to maintain the population diversity and avoid prematurity, each q-bit string of individual provides its evolutionary history information to another. To demonstrate its effectiveness and applicability, the proposed algorithm was tested on two famous combinatorial optimization problems, namely, the knapsack problem and flow shop problem. The results show that MMQEA performs very well compared to quantum evolutionary algorithm (QEA) and the conventional genetic algorithm. © 2012 Institute of Electrical Engineers of Japan.

Multiobjective process planning and scheduling (PPS) is a most important practical but very intractable combinatorial optimization problem in manufacturing systems. Many researchers have used multiobjective evolutionary algorithms (moEAs) to solve such problems
however, these approaches could not achieve satisfactory results in both efficacy (quality, i.e., convergence and distribution) and efficiency (speed). As classical moEAs, nondominated sorting genetic algorithm II (NSGA-II) and SPEA2 can get good efficacy but need much CPU time. Vector evaluated genetic algorithm (VEGA) also cannot be applied owing to its poor efficacy. This paper proposes an improved VEGA with archive (iVEGA-A) to deal with multiobjective PPS problems, with consideration being given to the minimization of both makespan and machine workload variation. The proposed method tactfully combines the mechanism of VEGA with a preference for the edge region of the Pareto front and the characteristics of generalized Pareto-based scale-independent fitness function (gp-siff) with the tendency to converge toward the central area of the Pareto front. These two mechanisms not only preserve the convergence rate but also guarantee better distribution performance. Moreover, some problem-dependent crossover, mutation, and local search methods are used to improve the performance of the algorithm. Complete numerical comparisons show that the iVEGA-A is obviously better than VEGA in efficacy, and the convergence performance is also better than NSGA-II and SPEA2, while the distribution performance is comparable to and the efficiency is obviously better than theirs. © 2012 Institute of Electrical Engineers of Japan.

This paper proposes a parallel quantum evolutionary algorithm (PQEA) using Client-Server model for multi-objective optimization problems. Firstly, the PQEA uniformly decomposes a multi-objective optimization problem into a number of scalar optimization sub-problems. All the sub-problems are classified into several groups according to their similarities. Each "Client" processes the evolution for a group of neighbor sub-problems in parallel. There is a quantum individual used to address the sub-problems of a group in a "Client". Since the quantum individual is a probabilistic representation, it can share evolutionary information of the neighbor sub-problems in one group, while the sub-problems are orderly solved using a same q-bit individual. The "Server" maintains non-dominated solutions that are generated by every "Client". The current best solution for each sub-problem can be found in the "Server", when the quantum individual updated its states for evolution. Experimental results have demonstrated that PQEA obviously outperforms the most famous multi-objective optimization algorithms MOEA/D on the bi-objectives. For the more objectives, the PQEA obtains the similar results with MOEA/D, even with the same evaluation times. Furthermore, in this paper, the scalability and sensitivity of PQEA have also been experimentally investigated.

The unsmooth job flow between the production shops along a manufacturing chain is a problem commonly seen in industries due to the inconsistency in processing speed and delivery between the chained production shops. With many factors involved in the coordination between production shops, the problem is complex and few solutions have been provided to date. This study presents an approach able to solve this problem by implementing time-based manufacturing that enables the speed and timing of each shop's job outflow to match those of its successor shop's job inflow. The proposed method is composed of offline schedule making and online job processing control. It aims to complete each job in a just-in-time (JIT) manner at the time the job is wanted by the next production shop. Designed upon a flexible job shop environment, which is easy to be transformed into other shops with similar characters, the proposed method is expected to be widely applicable to JIT scheduling problems. An industrial case study is made and results show that the proposed method has a strong ability in JIT job completion, tardy job prevention and makespan reduction.

This paper considers the dynamic flexible-flow shop scheduling problem in which the manufacturing environment consists of multiple stages with multiple machines at each stage. All jobs must go through all stages in sequence in order to become a product, and the processing time for each product on each machine is different. There exists a delivery time between machines of neighboring stages, and the release time of each job is unknown, which means that a new job may be released into the system at any time. The objective of this work is to minimize the total earliness and tardiness penalties of all jobs, or to achieve just-in-time production. Previous researches did much on the static scheduling of such problem with different objectives, whereas little effort has been made on dynamic scheduling for such problem, which is more difficult than the static problem but becomes more and more important under the increasingly competitive manufacturing industry. Therefore, to address such need, efforts are been made on dynamic scheduling and several distributed-intelligence approaches are proposed in this paper. Featured by concurrent computing, the distributed-intelligence approaches have been tested on two different manufacturing environments, whose result tells that the proposed approaches deliver competitive performance for the targeted problem. (C) 2010 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.

when the demand of system productivity is increased, the throughput of AGV system is also needed to be developed. By adding more single-load AGVs in the system, more items can be transported at same time. But consequently, the traffic for AGVs becomes more congested, and tradeoff caused by number of AGV is difficult to balance. The multi-load AGV can transport more items at same time and no additional AGVs are necessarily involved. But the control of multi-load AGV is much more complicated than single-load one as realized as many aspects. In this paper, a simulation study of multi-load AGV system based on JIT production environment is presented, the purpose of which is to examine the improvement generated by multi-load AGV system in JIT production environment and the performance of it when higher demand of system productivity is met. The control of multi-load AGV system focuses on selection of requesting stations and items in different situations, and also the responding procedure. The results of experimental test indicate that the introduced multi-load AGV system is adaptive to JIT production system and able to provide much higher throughput to satisfy the increasing demand of productivity compared with single-load one. © 2010 IEEE.

This paper considers the dynamic flexible flow shop problem with unrelated parallel machines at each stage and the objective of the problem is to minimize the total earliness and tardiness penalties of all jobs, or to achieve weighted just in time. In previous studies, the job will never change if it has been assigned to a certain machine. Since the schedule should be updated according to the change of current stages, this paper proposes a mechanism that allows the selected machines to be updated for better schedule. Two kinds of arriving time control algorithms are developed. In addition, three improved ways are used to enhance the efficiency of the two algorithms. At last, these eight different approaches have been compared with two classic methods - the stage to stage feedback approach and the earliest to delivery dispatching rule and computational experiences reveal that the proposed approaches are efficient for the targeted problem. © 2010 IEEE.

Scheduling for modern professional sports leagues has drawn considerable attention in recent years in that their practical applications involve significant revenues and generate challenging combinatorial optimization problems. The Traveling Tournament Problem is a sports scheduling problem that abstracts the important issues in creating time tables: feasibility and team travel, where the objective is to minimize the total distance traveled by the teams. In this paper, we tackle the mirrored version of this problem. First, an effective and comprehensive constructive algorithm is applied which quickly obtains initial solution at a very high quality. Then a hybrid local search approach was proposed based on the combination of Tabu Search and Variable Neighborhood Descent meta-heuristic, together with Greedy Randomized Adaptive Search Procedure, which explores large neighborhood with various and effective moves. Very competitive solutions are obtained for benchmark instances within a reasonable amount of time compared with previous results in the literature. © 2010 IEEE.

The time-cost trade-off problem (TCTP) is an important branch in the project scheduling problem. However, the duration and cost of each activity could change stochastically as a result of uncertain factors. To meet the needs of real projects, an improved approach based on trapezoid fuzzy numbers is applied to estimate the uncertainty of time and cost. And then α-cut method is applied to decide the risk level. Furthermore, improved crossover and mutation methods for multi-objective genetic algorithm (MOGA) are used to make a large-scale computation possible. The efficiency of the proposed approach is verified by comparison with previous researches. In addition, economic analysis skill of finance cost is integrated into the new model to provide greater flexibility to managers when making decisions. Finally, time-cost tables under different risk levels for case examples are given and the advantages are investigated based on computation results. © 2010 IEEE.

With the high speed innovation of information technology, many production scheduling systems have been developed. However, a lot of customization according to individual production environment is required, and then a large investment for development and maintenance is indispensable. Therefore now the direction to construct scheduling systems should be changed. The final objective of this research aims at developing a system which is built by it extracting the scheduling technique automatically through the daily production scheduling work, so that an investment will be reduced. This extraction mechanism should be applied for various production processes for interoperability. Using the master information extracted by the system, production scheduling operators can be supported to accelerate the production scheduling work easily and accurately without any restriction of scheduling operations. By installing this extraction mechanism, it is easy to introduce a scheduling system without a lot of expense for customization. In this paper, first a model for expressing a scheduling problem is proposed. Then the guideline to extract the scheduling information and use the extracted information is shown and some applied functions are also proposed based on it. (C) 2009 Wiley Periodicals, Inc. Electron Comm Jpn, 93(1): 19-29, 2010; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/ecj.10188

With the high speed innovation of information technology, many production scheduling systems have been developed. However, a lot of customization according to individual production environment is required, and then a large investment for development and maintenance is indispensable. Therefore now the direction to construct scheduling systems should be changed. The final objective of this research aims at developing a system which is built by it extracting the scheduling technique automatically through the daily production scheduling work, so that an investment will be reduced. This extraction mechanism should be applied for various production processes for interoperability. Using the master information extracted by the system, production scheduling operators can be supported to accelerate the production scheduling work easily and accurately without any restriction of scheduling operations. By installing this extraction mechanism, it is easy to introduce a scheduling system without a lot of expense for customization. In this paper, first a model for expressing a scheduling problem is proposed. Then the guideline to extract the scheduling information and use the extracted information is shown and some applied functions are also proposed based on it. (C) 2009 Wiley Periodicals, Inc. Electron Comm Jpn, 93(1): 19-29, 2010; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/ecj.10188

This paper proposed a new evolutionary algorithm based on concept and principles of quantum computing, called Multi-update Mode Quantum Evolution Algorithm (MMQEA), in which having two update modes beta-update and beta-update, and between their modes each update procedure provide its evolutionary information to other one, that to guide the other update-mode to maintain the population diversity and avoid premature. Meanwhile in MMQEA, proposed a new individual structure that composed by two Q-bit strings. Applying the multi-update mode to individual evolution improved the Q-gate updating efficiency in each generation of MMQEA. To demonstrate its effectiveness and applicability, the proposed algorithms were tested on a famous combinatorial optimization problem, the knapsack problem. The results show that MMQEA performs very well compared with Quantum Evolutionary Algorithm (QEA).

In the practice of scheduling of construction projects, there is a great variety of methods and procedures that need to be selected at each construction process during project. Accordingly, it is important to consider the different modes that may be selected for an activity in the scheduling of construction projects. In this study, first, we mathematically formulate the resource-constrained project scheduling problem with multiple modes while minimizing the total project time as the objective function. Following, we propose a new random key-based genetic algorithm approach which includes the mode reduction procedures to solve this NP-hard optimization problem. Finally, in order to evaluate the performance of our method, we are scheduled in the close future to implement the proposed approach on some standard project instances as the computational experiment and analyze these experimental results comparing with the bi-population-based genetic algorithm by Peteghem and Vanhoucke [1].

This paper presents some important improvements to a previously proposed intelligent production system dealing with a dynamic flexible flow shop scheduling problem under a multi-stage multi-machine factory environment. These improvements greatly help upgrade the overall system performance under stable demand situations as well as under fluctuated demand situations, build the system robust against demand increase, and raise the systems machine utilization rate. The research objective is to minimize the total earliness and tardiness penalties of all jobs during any given period of time. The system works on the basis of multi-agent feedbacks that are conducted by agents which collect realtime information, make decisions, and work interactively to give corresponding solutions to each job. Comparison between the previous system and the improved one has been carried out, and the experimental results demonstrate the effectiveness of the proposed improvements under various system situations. © 2010 IEEE.

The process planning and scheduling (PPS) is to determine a solution (schedule), which tells a production facility what to make, when, and on which equipment, to process a set of parts with operations effectively. Multiobjective PPS problems become more complex because the decision maker need to make a trade-off between two or more objectives while determining a set of optimal nondominated solutions effectively. The previous research works use evolutionary algorithms (EA) to solve such problems, however, the proposed approaches cannot get a good balance between efficacy and efficiency. This paper proposed an improved vector evaluated genetic algorithm with archive (iVEGA-A) mechanism to deal with PPS problem while considering the minimization of the makespan and minimization of the variation of workload of machine. The proposed algorithm has been compared with other approaches to verify and benchmark the optimization reliability on PPS problems. These comparisons indicate iVEGA-A is better than vector evaluated genetic algorithm (VEGA) did on efficacy and negligible difference on efficiency. The efficacy is not less than some famous approaches, such as, nondominated sorting genetic algorithm II (NSGA-II) and strength Pareto evolutionary algorithm 2 (SPEA2) and the efficiency is obviously better than the latter. ©2010 IEEE.

Weighted linear scalar, which transfers a multi-objective problem to many single objective sub-problems, is a basic strategy in traditional multi-objective optimization. However, it is not well used in many multi-objective evolutionary algorithms because of most of them are lack of balancing between exploitation and exploration for all sub-problems. This paper proposes a novel multi-objective evolutionary algorithm called multi-objective quantum evolutionary algorithm (MOQEA). Quantum evolutionary algorithm is a recent developed heuristic algorithm, based on the concept of quantum computing. The most merit of QEA is that it has little q-bit individuals are evolved to obtain an acceptable result. MOQEA decomposes a multi-objective optimization problem into a number of scalar optimization sub-problems and optimizes them simultaneously. Each sub-problem is optimized by one q-bit individual. The neighboring solutions that are defined as a set of nondominated solutions of sub-problem are generated from the corresponding q-bit individual. The experimental results have demonstrated that MOQEA outperforms or performs similarly to MOGLS and NSGA-II on discrete multi-objective problems.

The process planning and scheduling (PPS) is to determine a solution (schedule), which tells a production facility what to make, when, and on which equipment, to process a set of parts with operations effectively. Multiobjective PPS problems become more complex because the decision maker need to make a trade-off between two or more objectives while determining a set of optimal nondominated solutions effectively. The previous research works use evolutionary algorithms (EA) to solve such problems, however, the proposed approaches cannot get a good balance between efficacy and efficiency. This paper proposed an improved vector evaluated genetic algorithm with archive (iVEGA-A) mechanism to deal with PPS problem while considering the minimization of the makespan and minimization of the variation of workload of machine. The proposed algorithm has been compared with other approaches to verify and benchmark the optimization reliability on PPS problems. These comparisons indicate iVEGA-A is better than vector evaluated genetic algorithm (VEGA) did on efficacy and negligible difference on efficiency. The efficacy is not less than some famous approaches, such as, nondominated sorting genetic algorithm II (NSGA-II) and strength Pareto evolutionary algorithm 2 (SPEA2) and the efficiency is obviously better than the latter. ©2010 IEEE.

This research considers the just-in-time shceduling problem aiming at minimizing the total earliness and tardiness penalties of all jobs. The manufacturing environment consists of multiple stages with multiple machines at each stage. All jobs must go through all the stages in sequence in order to become a product, and the processing time for each product on each machine is different. There exist delivery time between machines, and the release time of jobs is unknown, which means a new job may be released into the sysyem at any time. In addition, machine breakdown and machine recovery are also taken into consideration. An intelligent production control system is presented in this paper to solve the described problem. The system, consisted of distributed agents, is featured by concurrent computing and realtime communication. The system has been tested under normal manufacturing condition as well as machine breakdown situation. The result reveals that the proposed intelligent production control system delivers competitive performance for the targeted problem.

This paper proposed the real-time buffer management method which extends the concept of 'buffer' by using Drum-Buffer-Rope (DBR) scheduling technique of Theory of Constraints (TOC) according to real-time information. The current DBR buffer is used to supervise ascertain whether the product in the last stage of the production process, stages managed by the buffer, has arrived. If the product has not arrived the warning sign continues to be emitted and the check action for that situation is left to the operators. The problem arises when a meaningless warning sign is emitted for too long and it affects the operators. The mechanism proposed in this paper suppresses the meaningless warnings by using the detailed schedule information provided by the on-site operators and actual result information obtained by the on-site instruments. Copyright © 2009 Inderscience Enterprises Ltd.

Worker allocation problem of the manual labor factory consists in deciding who does what during the manual labor production. Manual labor factory, has concern for reducing the production duration, human cost and overwork. A novel scheduling model of worker allocation problem for manual labor is proposed in this paper. Given a set of jobs and a set of workers, worker allocation problem is to assign the workers into these jobs to reduce human cost, to shorten production duration and control production overwork. Worker allocation problem is a very important issue in the production scheduling, since total budget and human resource involved must be managed optimally. Worker allocation problem is NP-Hard problem to find optimal solution as increasing worker number and enlarging production scale. Genetic algorithm (GA) is considered a promising method so in this paper worker allocation problem is solved by using genetic algorithm. This paper shows it is very flexible and accurate for finding near-optimal solutions in the short time.

The flexible flow shop refers to such a manufacturing environment in which jobs are to be processed through serial stages, with one or multiple machines available at each stage. It is usually a complex task when specific objective is demanded such as minimum cost, minimum time, etc. Static scheduling of such problems has been much researched, however, little efforts have been made on real-time scheduling when the release time of each job is unknown. In this paper, some online scheduling methods are presented to deal with the tough problem of real-time Just-In-Time manufacturing. In addition to applicable dispatching rules, agent-based approaches are also proposed featuring feedback learning, and realtime prediction. The simulation result reveals that the presented distributed learning approach, especially when combined with realtime prediction, delivers a high performance.

The problem considered in this research is the just-in-time scheduling of a manufacturing environment that is able to produce several different products. New jobs come randomly into the system, expected to become one of the products. Each job must go through multiple stages before it can be finished as a product There are multiple machines at each stage, and the processing time of each product on each machine is different There exits delivery time between stages. Previous researches did much on the static scheduling of such problem by using mixed integer linear programming. However, little efforts have been made on realtime scheduling, which means the release time of jobs is unknown. But such scheduling is becoming more and more important under the increasingly competitive manufacturing market In this paper, two distributed feedback mechanisms are proposed to solve the realtime scheduling problem of minimizing earliness and tardiness penalties of all jobs. The simulation shows that the proposed distributed feedback mechanisms deliver quite competitive performance for the targeted problem.

Earliness and tardiness penalties are designed for Such scheduling problems where the popular Just-In-Time (JIT) concept is considered to be of significant importance. In this paper, a self evolution (SE) algorithm is proposed to solve the problem of single-machine total earliness and tardiness penalties with a common due date. Up to now, no specific attention has been paid to straddling V-shaped schedules of such problems, which may be better than pure V-shaped schedules for the early due date cases; and no specific discussions have been made on the start time setting of the first job in a schedule. Therefore, in this research, efforts have been made on digging out the straddling V-shaped schedules, improving the efficiency of setting the start time of a schedule, and reducing the execution time. In addition, a new RHRM approach is proposed to create the initial solution for evolution, which helps in achieving the fast contingency of the algorithm. The performance of the proposed algorithm has been tested on 280 benchmark instances ranging from 10 to 1000 jobs from the OR Library, and the results show that the proposed SE algorithm delivers much higher efficiency in finding optimal or near-optimal solutions with both better results in total penalties and significant execution time reduction. (C) 2008 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.

With the high speed innovation of information technology, many production scheduling systems have been developed. However, a lot of customization according to individual production environment is required, and then a large investment for development and maintenance is indispensable. Therefore now the direction to construct scheduling systems should be changed. The final objective of this research aims at developing a system which is built by it extracting the scheduling technique automatically through the daily production scheduling work, so that an investment will be reduced. This extraction mechanism should be applied for various production processes for the interoperability. Using the master information extracted by the system, production scheduling operators can be supported to accelerate the production scheduling work easily and accurately without any restriction of scheduling operations. By installing this extraction mechanism, it is easy to introduce scheduling system without a lot of expense for customization. In this paper, at first a model for expressing a scheduling problem is proposed. Then the guideline to extract the scheduling information and use the extracted information is shown and some applied functions are also proposed based on it.

The Just-In-Time (JIT) concept is of great importance in many manufacturing processes. JIT scheduling problems affects the performance of the whole production procedure, because early in job completion causes inventory cost while delay in job completion raises penalties paid to customers. In this paper, a memorial self evolution algorithm is proposed to solve the problem of total earliness and tardiness penalties on a machine unit with a common due date. Up to now, researches on this problem have paid no specific attention to straddling V-shaped schedules, which may be better than pure V-shaped schedules for early due date cases; and no specific discussions have been made on the start time setting of the first job in a schedule. Thus, efforts have been made on searching good straddling V-shaped schedules, and optimizing start time setting of schedules. A GHRM approach is proposed to create the initial solution for memorial self evolution. Meanwhile a database which keeps the memories of the elite solutions is introduced to deliver better initial solutions for similar problems. The performance of the proposed algorithm has been tested on 280 benchmark instances ranging from 10 to 1000 jobs. The results show that the proposed memorial self evolution algorithm delivers better results in in finding optimal or near-optimal solutions than previous researches.

The Just-In-Time (JIT) concept is of great importance in many manufacturing processes. JIT scheduling problems affects the performance of the whole production procedure, because early in job completion causes inventory cost while delay in job completion raises penalties paid to customers. In this paper, a memorial self evolution algorithm is proposed to solve the problem of total earliness and tardiness penalties on a machine unit with a common due date. Up to now, researches on this problem have paid no specific attention to straddling V-shaped schedules, which may be better than pure V-shaped schedules for early due date cases
and no specific discussions have been made on the start time setting of the first job in a schedule. Thus, efforts have been made on searching good straddling V-shaped schedules, and optimizing start time setting of schedules. A GHRM approach is proposed to create the initial solution for memorial self evolution. Meanwhile a database which keeps the memories of the elite solutions is introduced to deliver better initial solutions for similar problems. The performance of the proposed algorithm has been tested on 280 benchmark instances ranging from 10 to 1000 jobs. The results show that the proposed memorial self evolution algorithm delivers better results in in finding optimal or near-optimal solutions than previous researches.©2008 IEEE.

Inventory cost and delay penalty are two kinds of annoying spendings in manufactory industry. Accordingly, earliness and tardiness penalties are proposed to simulate such scheduling problems where the popular just-in-time (JIT) concept is considered to be of significant importance. In this paper, a self evolution algorithm is proposed to solve the problem of single machine total earliness and tardiness penalties with a common due date. Up to now, such problem has been solved without specific consideration of straddling V-shaped schedules, which may be better than pure V-shaped schedules for early due date problems; without specific discussions on g improving, where g refers to the idle time before the start of the first job; and the many individuals in all so far proposed GA-like algorithms become the bottleneck of execution time reduction. Therefore, in this research, efforts have been made on digging out the straddling V-shaped schedules, improving the efficiency of g improving, and reducing the execution time. In addition, a new RHRM approach is proposed to create the initial solution for evolution, which helps achieve the fast contingency of the algorithm. The performance of the proposed algorithm has been tested on 280 benchmark instances ranging from 10 to 1000 jobs from the OR Library, the results showing that the proposed self evolution algorithm delivers much higher efficiency in finding optimal or near-optimal solutions with both better results in total penalties and significant execution time reduction.

There are many kinds of delay in real-world production systems caused by many reasons including unexpected accidents. A delay of order may inflict great damages for not only itself but also the other affected orders. To prevent these types of loss from frequent delay, DBR (Drum-Buffer-Rope) scheduling method of TOC (Theory of Constraints) manages production schedule observing the state of time buffers. The current buffer size setting method for DBR scheduling is very simple and depends on user's experience. Although it makes possible to keep the due time for production orders, it leads to the redundant production lead time and stock. For DBR scheduling, it is not clear how the buffer size should be set. Therefore, this paper proposes a buffer size setting method for DBR scheduling providing a numerical model for the buffer size. In addition, a simulation gives the result of comparison between the current method and the proposed method, and the effect of the proposed method is shown.

With the high speed innovation of information technology, many production scheduling systems (to support scheduling work) have been developed. However, a significant customization according to individual production environment is required, and then a large investment for development and maintenance is indispensable. Therefore now the direction to construct scheduling systems should be changed. The final objective of this research aims at developing a system which is built by itself extracting the scheduling technique automatically through the daily production scheduling work. By using the information extracted by the system, production scheduling operators can be supported to accelerate the production scheduling work easily and accurately without any restriction of scheduling works. In order to realize such a production scheduling system, the mechanism to extract scheduling information is needed. In this paper, using Gantt Chart Interface System that is developed to emulate the production scheduling work on papers through the analysis of operator's works, the method extracting and generalizing the relation information between operations from user manipulation is proposed. And the operator assistant function which uses the extracted information is proposed. Then by conducting experiment that uses the assistant function, the availability of the method of extracting and generalizing is shown.

Although decomposition procedure is a common approach to handle computational complexity which is caused by large scale production scheduling problems, in order to extend the applicability of the decomposition approach, the overall problem should be decomposed on which criterion is always a problem. Also how to improve decomposition approach's computational efficiency and resultant effectiveness is still a challenging task in this field.
With the purpose of improving decomposition approach's efficiency and decreasing computational complexity for job shop scheduling problems, a new dynamic local focusing approach which has a machine-based divided criterion focused on the longest active chain is proposed in this paper. Not liking the used decomposition procedure, the proposed one does not decompose a job shop into cells at the very beginning. It dynamically classifies the machines, which process operations on the identified longest active chain of the whole problem as one focused cell (decomposed sub-problem). The schedule is improved by redefining, adjusting and solving the focused cell schedule which is updated iteratively with the entire schedule's longest active chain in this dynamic procedure. The proposed approach is tested on make-span minimum benchmark job shop scheduling problems. Test results show that the algorithm is capable of efficiently generating good schedules.

This paper deals with the partial scheduling adjustment in the autonomous distributed scheduling method. The scheduling adjustment for the production system is required when uncertainty occurs. In that case, it does not need to adjust the whole schedule, but only the partial schedule which receives a direct influence by uncertainty according to the variable situation. From the viewpoint of the distributed system, to implement a function which adjusts such a partial schedule, each machine needs to manage schedule and exchange information. It has been studied by some researchers in recent years. However, the partial scheduling adjustment has not been controlled to change the processing sequence of operations in a machine, namely interchanging operations. Therefore, we propose realizing a function interchanging operations in a machine to diminish the influence occurred by a delay of a certain operation.

This paper deals with the partial scheduling adjustment in the autonomous distributed scheduling method. The scheduling adjustment for the production system is required when uncertainty occurs. In that case, it does not need to adjust the whole schedule, but only the partial schedule which receives a direct influence by uncertainty according to the variable situation. From the viewpoint of the distributed system, to implement a function which adjusts such a partial schedule, each machine needs to manage schedule and exchange information. It has been studied by some researchers in recent years. However, the partial scheduling adjustment has not been controlled to change the processing sequence of operations in a machine, namely interchanging operations. Therefore, we propose realizing a function interchanging operations in a machine to diminish the influence occurred by a delay of a certain operation.

At present, many production scheduling systems have been developed. However, a lot of customization according to individual production environment is required, and then a large investment for development and maintenance is indispensable. Therefore now the direction to construct scheduling systems should be changed. The final objective of this research aims at developing a system which is built by it extracting the scheduling technique automatically through the daily production scheduling work, so that an investment will be decreased. This paper shows a prototype of Gantt Chart Interface System that emulates the production scheduling work on papers. Through the analysis of operator's operations, this paper proposes how to extract the master information required by production scheduling. Using the information, the scheduling operators can be supported to accelerate the scheduling work easily and accurately without any restriction of scheduling operations. By installing this extraction mechanism, it is easy to introduce scheduling system without a lot of expense for customization. In this paper, the extraction mechanism of the master information and operator support functions are proposed through a model expressing a scheduling problem.

Ant Colony Optimization (ACO) is recently proposed as one of meta-heuristic methods for solving, hard combinatorial optimization problems.,,Job-shop Scheduling Problem (JSP) is to determine the, operation sequences on the machines for minimizing a makespan, and also has been confirmed as one of NP-hard problems.
In this paper, an ACO approach for solving JSP is presented. We propose hybrid ACO that is combined other heuristic method to find good solution. The traditional heuristic methods include shortest processing time (SPT), longest processing time (LPT), shortest remaining time (SRT), longest remaining time (LRT), longest remaining time excluding the operation under consideration (LRM) for assigning priorities to the unscheduled operations is defined. Some numerical examples are demonstrated to show the performance of JSP using hybrid ACO.

資料形態 : テキストデータ プレーンテキスト

資料形態 : テキストデータ プレーンテキスト

The paper describes attempts to provide a shell for building intelligent systems in consideration of a tight integration of the knowledge representation, the inference mechanism, the methodology of construction and the development environment. The Shell is called HyperAuk: Shell for building Intelligent Systems. It is a knowledge engineer's assistant tool that provides the graphical development environment to support the construction process, that includes planning, editing, debugging and testing phases.<br>HyperAuk enhances the object oriented knowledge representation auk: autonomic knowledge representation, we proposed. The methodology behind HyperAuk provides the concept of the top-down planning, the bottom-up construction and the phaseless development, and defines the sequence of its construction process. According to this concept, HyperAuk constructs the development environment to support the whole development process. This environment is also including the function of the graphical user interface construction. HyperAuk provides a uniform framework by generalization of top-down planning from the methodology view. So the environment for top-down planning is able to be used in any cases.<br>HyperAuk is now operational using Smalltalk-80, and its environment is constructed using the user interface management system Talkie, we also developed.

This paper describes user friendly construction methods for batch plant scheduling systems. Many kinds of flexibility handling many sorts, low quantity and high value products are indispensable for a plant operating scheduling system. Sorts of products are changed little by little in demand. For that, the plant may be reconfigured. In such case, the scheduling system must have an ability to customize easily by plant operators. We have already proposed a framework for an actual plant operating scheduling system. As the enhancement of this system, this paper describes a definition method for the automatic scheduling strategies based on the visual programming technology and a manual modification method for a plan that is scheduled by the automatic scheduling mechanism. This System is implemented on Objectworks_??_Smalltalk. It is applied on many actual plants, so in this paper an example is also described.

記事分類: 管理工学

本稿では 知的システム構築用シェルAUKについて報告する.AUKは 効率のよい知的システムの開発を目的とし 知識ベースの構築 および開発されたシステムの利用において 優れた操作性を実現するという観点から設計を行ったシェルである.このシェルの開発にあたって まず オブジェクト指向知撤表現auk(autonomic knowledge unit:自律的知識単位)を提案した.この知識表現では 知識をオブジェクトとして扱い そのオブジェクトの外部インタフェースを定め 知識のモジュール性 再利用性 統合性を高めた.また 種々の表現に対応すべく 拡張が容易であるように設計した このような知識表現をベースに 知的システム構築用シェルAUKを Smalltalk-80上に開発した.AUKにおける知識ベースの構築は この知識表現の特徴を生かし グラフィカルな知識ベース開発環鏡内で インタラクティブに進めることができ 効率的な開発が可能である.