Often in manufacturing systems, scenarios arise where the demand for maintenance exceeds the capacity of maintenance resources. This results in the problem of allocating the limited resources among machines competing for them. This maintenance scheduling problem can be formulated as a Markov decision process (MDP) with the goal of finding the optimal dynamic maintenance action given the current system state. However, as the system becomes more complex, solving an MDP suffers from the curse of dimensionality. To overcome this issue, we propose a two-stage approach that first optimizes a static condition-based maintenance (CBM) policy using a genetic algorithm (GA) and then improves the policy online via Monte Carlo tree search (MCTS). The static policy significantly reduces the state space of the online problem by allowing us to ignore machines that are not sufficiently degraded. Furthermore, we formulate MCTS to seek a maintenance schedule that maximizes the long-term production volume of the system to reconcile the conflict between maintenance and production objectives. We demonstrate that the resulting online policy is an improvement over the static CBM policy found by GA. Note to Practitioners—: This article proposes a method of scheduling maintenance in complex manufacturing systems in scenarios where there is frequent competition for maintenance resources. We use a condition-based maintenance policy that prescribes maintenance actions based on a machine's current health. However, when several machines are due for maintenance, a maintenance technician must choose between multiple competing jobs. While a common approach is to establish rules that dictate how maintenance jobs should be prioritized, such as the first-in, first-out rule, the goal of this work is to improve upon static policies in real time. We do this by strategically evaluating sequences of maintenance actions and playing out many "what-if" scenarios to see how the system will behave in the future. Implementation of the proposed method relies on the construction of a simulation model of the target system. This model is capable of retrieving the current state of the physical system, including the degradation state of machines, the availability of maintenance resources, and the distribution of parts throughout buffers in the system. We present several simulation experiments that demonstrate the improvement in system performance that our approach provides. Future work will aim to improve the efficiency of maintenance prioritization through online learning as well as more accurately identify manufacturing system configurations that will yield the greatest benefit of these methods.

Quality is a key determinant in deploying new processes, products, or services and influences the adoption of emerging manufacturing technologies. The advent of additive manufacturing (AM) as a manufacturing process has the potential to revolutionize a host of enterprise-related functions from production to the supply chain. The unprecedented level of design flexibility and expanded functionality offered by AM, coupled with greatly reduced lead times, can potentially pave the way for mass customization. However, widespread application of AM is currently hampered by technical challenges in process repeatability and quality management. The breakthrough effect of six sigma (6S) has been demonstrated in traditional manufacturing industries (e.g., semiconductor and automotive industries) in the context of quality planning, control, and improvement through the intensive use of data, statistics, and optimization. 6S entails a data-driven DMAIC methodology of five steps-define, measure, analyze, improve, and control. Notwithstanding the sustained successes of the 6S knowledge body in a variety of established industries ranging from manufacturing, healthcare, logistics, and beyond, there is a dearth of concentrated application of 6S quality management approaches in the context of AM. In this article, we propose to design, develop, and implement the new DMAIC methodology for the 6S quality management of AM. First, we define the specific quality challenges arising from AM layerwise fabrication and mass customization (even one-of-a-kind production). Second, we present a review of AM metrology and sensing techniques, from materials through design, process, and environment, to postbuild inspection. Third, we contextualize a framework for realizing the full potential of data from AM systems and emphasize the need for analytical methods and tools. We propose and delineate the utility of new data-driven analytical methods, including deep learning, machine learning, and network science, to characterize and model the interrelationships between engineering design, machine setting, process variability, and final build quality. Fourth, we present the methodologies of ontology analytics, design of experiments (DOE), and simulation analysis for AM system improvements. In closing, new process control approaches are discussed to optimize the action plans, once an anomaly is detected, with specific consideration of lead time and energy consumption. We posit that this work will catalyze more in-depth investigations and multidisciplinary research efforts to accelerate the application of 6S quality management in AM.

Patient satisfaction is a key performance indicator of patient-centered care and hospital reimbursement. To discover the major factors that affect patient experiences is considered as an effective way to formulate corrective actions. A patient during his/her healthcare journey interacts with multiple health professionals across different service units. The health-related data generated at each step of the journey is a valuable resource for extracting actionable insights. In particular, self-reported satisfaction survey and the associated patient electronic health records play an important role in the hospital-patient interaction analysis. In this paper, we propose an interpretable machine learning framework to formulate the patient satisfaction problem as a supervised learning task and utilize a mixed-integer programming model to identify the most influential factors. The proposed framework transforms heterogeneous data into human-understandable features and integrates feature transformation, variable selection, and coefficient learning into the optimization process. Therefore, it can achieve desirable model performance while maintaining excellent model interpretability, which paves the way for successful real-world applications.

Identifying drug-drug interactions (DDIs) is a critical enabler for reducing adverse drug events and improving patient safety. Generating proper DDI alerts during prescribing workflow has the potential to prevent DDI-related adverse events. However, the implementation of DDI alerting system remains a challenge as users are experiencing alert overload which causes alert fatigue. One strategy to optimize the current system is to establish a list of high-priority DDIs for alerting purposes, though it is a resource-intensive task. In this study, we propose a machine learning framework to extract useful features from the FDA adverse event reports and then identify potential high-priority DDIs using an autoencoder-based semi-supervised learning algorithm. The experimental results demonstrate the effectiveness of using adverse event feature representations in differentiating high- and low-priority DDIs. Additionally, the proposed algorithm utilizes stacked autoencoders and weighted support vector machine for boosting classification performance, which outperforms other competing methods in terms of F-measure and AUC score. This framework integrates multiple information sources, leverages domain knowledge and clinical evidence, and provides a practical approach for pre-screening high-priority DDI candidates for medication alerts.

In the personalized patient-centered healthcare, self-reported patient satisfaction survey data plays an important role. Given the patient survey data, it is necessary to identify the drivers of patient satisfaction and explain them so that such patterns can be used in future as well as necessary corrective actions can be taken. In healthcare, both accuracy and interpretability are important criteria for choosing a reliable predictive model for analyzing patient data. Usually, complex models such as Random Forest, neural networks can achieve high prediction accuracy but lack necessary interpretation to their prediction results. In this paper, we address this problem by proposing a local explanation method to interpret complex model prediction results. First, we build a predictive model using Random Forest to fit the patient satisfaction data. Second, we utilize local explanation method to provide insights into the Random Forest prediction results so as to discover true reasons behind patient experiences and overall ratings. Specifically, our approach allows us to interpret patient's overall rating of a hospital at the individual level, and find out the set of the most influential factors for each patient. We focus on all unhappy patients to investigate the top reasons for patient dissatisfaction. Our approach and findings will help to establish guidelines for a quality healthcare.

The proliferation of real-time monitoring systems and the advent of Industrial Internet of Things (IIoT) over the past few years necessitates the development of scalable and parallel algorithms that help predict mechanical failures and remaining useful life of a manufacturing system or system components. Classical model-based prognostics require an in-depth physical understanding of the system of interest and oftentimes assume certain stochastic or random processes. To overcome the limitations of model-based methods, data-driven methods such as machine learning have been increasingly applied to prognostics and health management (PHM). While machine learning algorithms are able to build accurate predictive models, large volumes of training data are required. Consequently, machine learning techniques are not computationally efficient for data-driven PHM. The objective of this research is to create a novel approach for machinery prognostics using a cloud-based parallel machine learning algorithm. Specifically, one of the most popular machine learning algorithms (i.e., random forest) is applied to predict tool wear in dry milling operations. In addition, a parallel random forest algorithm is developed using the MapReduce framework and then implemented on the Amazon Elastic Compute Cloud. Experimental results have shown that the random forest algorithm can generate very accurate predictions. Moreover, significant speedup can be achieved by implementing the parallel random forest algorithm.

Community detection has attracted increasing attention during the past decade, and many algorithms have been proposed to find the underlying community structure in a given network. Many of these algorithms are based on modularity maximization, and these methods suffer from the resolution limit. In order to detect the underlying cluster structure, we propose a new convex formulation to decompose a partially observed adjacency matrix of a network into low-rank and sparse components. In such decomposition, the low-rank component encodes the cluster structure under certain assumptions. We also devise an alternating direction method of multipliers with increasing penalty sequence to solve this problem; and compare it with Louvain method, which maximizes the modularity, on some synthetic randomly generated networks. Numerical results show that our method outperforms Louvain method on the randomly generated networks when variance among cluster sizes increases. Moreover, empirical results also demonstrate that our formulation is indeed tighter than the robust PCA formulation, and is able to find the true clustering when the robust PCA formulation fails.

The last decade has witnessed an explosion in the modeling of complex systems. Predominantly, graphs are used to represent these systems. The problem of detecting overlapping clusters in graphs is of utmost importance. We present a novel definition of overlapping clusters. A noncooperative game is proposed such that the equilibrium conditions of the game correspond to the clusters in the graph. Several properties of the game are analyzed and exploited to show the existence of a pure Nash equilibrium (NE) and compute it effectively. We present two algorithms to compute NE and prove their convergence. Empirically, the running times of both algorithms are nearly linear in the number of edges. Also, one of the algorithms can be readily parallelized, making it scalable. Finally, our approach is compared with existing overlapping cluster detection algorithms and validated on several artificial and real data sets.

The selection and prioritization of research directions are always challenges. This paper aims to make sense of nanomaterial toxicity publication and keywords data through quantitative metrics and network visualization. We have adapted a combined approach of network analysis, co-occurrence analysis, clustering analysis and visual analytics, to characterize important relational properties of network structures and features of entities. The results show that both co-authorship network and keywords network on nanomaterial toxicity follow the power-law degree distribution. In addition, the co-authorship network appears to be of scale-free pattern. We also investigate and visualize the research trends in field of nanomaterial toxicity by studying top influence researchers and keywords over years. These findings offer researchers various insights of the patterns and trends in the nanomaterial toxicity.

NEIMiner is an integrated information system for studying the nanomaterial environmental impact (NEI). However, there is a lack of visual analytic tools that efficiently query and present large-scaled bibliography meta-data and NEI characterizations in a meaningful way. This paper presents the design and implementation efforts of developing the information visualization (InfoVis) module for NEIMiner. We first describe a user centered design approach to identify the analysis tasks, to select suitable visual representations, and to iteratively validate and improve the development. We then show that how existing techniques, such as graph simplification, enriched visualization algorithms and interactive features, can be usefully combined to aid users gaining insights. We demonstrate the utility of Info Vis through scenarios of constructing co-authorship network, bibliography keywords network, and nanomaterial terms co-occurrence network. We implement our techniques as a Drupal module. Our design is supportive for analysts and researchers to identify concepts and relationships in studying environmental impact of nanomaterial.

With the adoption of mobile, digital and social media networked crowds are reporting and acting upon events in smart environments. Existing platforms for crowdsourcing, support specific activity types, such as micro-tasks on the Amazon's Mechanical Turk; and fall short of facilitating general mechanisms for setting up and maintaining crowd networks easily, flexibly and in a variety of domains. Building upon First International Workshop on Ubiquitous Crowdsourcing, in this edition we challenge researchers and practitioners to identify requirements for a platform for crowd computing, arising from experiences in deployment crowdsourcing applications, which engage crowd members as sensors, controllers and actuators in smart cities and environments. This workshop brings together researchers to produce a vision for the universal crowdsourcing platform, documenting it in a theme publication.

There are several important properties of modern software systems. They tend to be large-scale with distributed and component-based architectures. Also, dynamic nature of operating environments leads them to utilize alternative algorithms. However, on the other hand, these properties make it hard to provide appropriate control mechanisms due to the increased complexity. Components are sharing resources and each component can have alternative algorithms. As a result, the behavior of a software system can be controlled through resource allocation, as well as algorithm selection. This novel control problem is worthy of investigation in order to double the benefits of those properties. In this paper, we design a control mechanism for such systems. The quality-of-service we are considering is a product of the value of solution and the time for generating solution for a given problem. We build a mathematical programming model that trades off these two conflicting objectives, and decentralize the model through an auction market. By periodically opening the auction market for each existing system state, a closed-loop policy is formed. Though similar problems can be found in multiprocessor scheduling literature, they have limitations in addressing this control problem. They commonly consider so-called workflow applications in which each component only has to process one task after all of its predecessors complete their tasks. In contrast, a component in the networks under consideration processes multiple tasks in parallel with its successors or predecessors.

Many organizations are expending enormous efforts to minimize cost and time when developing new products by sharing and reusing distributed design knowledge and information. Multiagent systems provide an ideal mechanism to develop various products efficiently and cost-effectively by integrating distributed design knowledge and information. Product family design entails developing a set of differentiated products around a platform, while targeting individual products to distinct market segments. This paper presents a mathematical model reflecting the tradeoffs inherent in having a platform and introduces a dynamic multiagent system (DMAS) based on negotiation mechanisms to address the model effectively. The negotiation protocol to design a platform is developed using market-based auction mechanisms. In the proposed system, specific design tasks are assigned to agents by decomposing tasks for product family design, and an optimal platform is determined by negotiations between agents. We apply the DMAS to design a platform for a family of power tools.

The main research focus of Web services is to achieve the interoperability between distributed and heterogeneous applications. Therefore, flexible composition of Web services to fulfill the given challenging requirements is one of the most important objectives in this research field. However, until now, service composition has been largely an error-prone and tedious process. Furthermore, as the number of available web services increases, finding the right Web services to satisfy the given goal becomes intractable. In this paper, toward these issues, we propose an AI planning-based framework that enables the automatic composition of Web services, and explore the following issues. First, we formulate the Web-service composition problem in terms of AI planning and network optimization problems to investigate its complexity in detail. Second, we analyze publicly available Web service sets using network analysis techniques. Third, we develop a novel Web-service benchmark tool called WSBen. Fourth, we develop a novel AI planning-based heuristic Web-service composition algorithm named WSPR. Finally, we conduct extensive experiments to verify WSPR against state-of-the-art AI planners. It is our hope that both WSPR and WSBen will provide useful insights for researchers to develop Web-service discovery and composition algorithms, and software.

A novel benchmark, WSBen, for testing web services discovery and composition is presented. WSBen includes: (1) A collection of synthetic web services (WSDL) files with diverse characteristics and sizes; (2) Test discovery and composition queries and solutions; and (3) External files for statistical analysis and AI planners. Users can fine-tune the generated WSDL files using various parameters such as skewness or matching type. It is our hope that WSBen will provide useful insights for researchers evaluating the performance of web services discovery and composition algorithms and softwares.

This paper presents a theoretical approach to determine the probability of misclassification of the multilayer perceptron (MLP) neural model, subject to weight errors. The type of applications considered are classification/recognition tasks involving binary input-output mappings. The analytical models are validated via simulation of a small illustrative example. The theoretical results, in agreement with simulation results, show that, for the example considered, Gaussian weight errors of standard deviation up to 22% of the weight value can be tolerated. The theoretical method developed here adds predictability to the fault tolerance capability of neural nets and shows that this capability is heavily dependent on the problem data.

We present a scheme incorporating neural network mappings for geometric modeling and interference determination in robotic collision detection. The scheme promises to greatly reduce the computational time associated with calculating collision points, which makes real-time obstacle avoidance more achievable. The scheme includes three modules: a geometric modeling module, a collision detection module, and a decision support module. The geometric modeling module employs the restricted Coulomb energy (RCE) paradigm to describe the spatial occupancy of a 3-D object by a number of overlapping spheres. The collision detection module receives the geometric pattern in the robot's environment and updates the spherical representation to perform geometric computation for existence of interference. The decision support module, using neural networks, provides online information for the collision detection module. A PUMA 560 robot's CAD model was built to test the scheme. The performances using the scheme and using the CAD model were compared and presented.

This paper presents the super relation graph (SRG) method for extracting prismatic features from the CAD boundary representation of a machined part. Using the definition of super relations and the validity of a feature volume, this method recognizes features with all three types of interactions: face splitting, face merging, and edge truncation.

This paper introduces an approach to cosmetic surface flaw identification that is essentially invariant to changes in workpiece orientation and position while being efficient in the use of computer memory. Visual binary images of workpieces are characterized according to the number of pixels in progressive subskeleton iterations. Those subskeletons are constructed using a modified Zhou skeleton transform with disk shaped structuring elements. Two coding schemes are proposed to record the pixel counts of succeeding subskeletons with and without lowpass filtering. The coded pixel counts are on-line fed to a supervised neural network that is previously trained by the backpropagation method using flawed and unflawed simulation patterns. The test workpiece is then identified as flawed or unflawed by comparing its coded pixel counts to associated training patterns. Such off-line trainings using simulated patterns avoid the problems of collecting flawed samples. Since both coding schemes tremendously reduce the representative skeleton image data, significant run time in each epoch is saved in the application of neural networks. Experimental results are reported using six different shapes of workpieces to corroborate the proposed approach.

This article presents the concept and development of a prototype diagnostic decision support system for real-time control and monitoring of dynamical processes. This decision support system, known as Diagnostic Evaluation and Corrective Action (DECA), employs qualitative reasoning, in conjunction with quantitative models, for monitoring and diagnosis of malfunctions in dynamical processes under routine operations and emergency situations. DECA is especially suited for application to time-constrained environments where an immediate action is needed to avoid catastrophic failure(s). DECA is written in common Lisp and has been implemented on a Symbolics 3670 machine; its efficacy has been verified using the data from the Three Mile Island No. 2 Nuclear Reactor Accident.