Abstract: To ensure undisrupted business, large Internet companies need to closely monitor various KPIs (e.g., Page Views, number of online users, and number of orders) of its Web applications, to accurately detect anomalies and trigger timely troubleshooting/mitigation. However, anomaly detection for these seasonal KPIs with various patterns and data quality has been a great challenge, especially without labels. In this paper, we proposed Donut, an unsupervised anomaly detection algorithm based on VAE. Thanks to a few of our key techniques, Donut greatly outperforms a state-of-arts supervised ensemble approach and a baseline VAE approach, and its best F-scores range from 0.75 to 0.9 for the studied KPIs from a top global Internet company. We come up with a novel KDE interpretation of reconstruction for Donut, making it the first VAE-based anomaly detection algorithm with solid theoretical explanation.

Abstract: The patient care processes in hospitals are supported by a range of operational activities including inventory management and distribution of supplies to point-of-care locations. Hospitals carry large amounts and a great variety of items, and the issues of storing and distributing these items throughout the hospital supply chain are of great importance to providing high-quality patient service. Healthcare logistics encompasses the process of handling physical goods (e.g. pharmaceuticals, surgical medical products, medical equipment, sterile items, linen, food, etc.) and the associated information flows, from the reception of the goods within a hospital to their delivery at patient care locations. The medical supply costs constitute the second largest expenditure in hospitals, after personnel costs. A high-performing supply chain may realize improved outcomes (e.g. safe and quality patient service) and greater efficiency. Logistics managers need to identify opportunities to improve the logistics processes in order to lower costs and to improve patient care quality. However, in order to improve the logistics processes, you must understand how the healthcare supply chain is currently performing. Measuring the performance of the supply chain is fundamental to identify and address deficiencies in the logistics activities, and it serves as a good input for managerial decision-making. The purpose of this article is to present existing research on performance measurement at the internal hospital supply chain (e.g. inventory management, distribution activities), and more specifically in the operating theatre since it is among the most critical resources for a hospital. At the operating theatre, the requested items should be available at the right time at the right place, in the right condition, at the lowest cost possible. Furthermore, we will also discuss literature on multi-criteria decision-making techniques. It enables researchers to build a performance measurement framework and to prioritize between multiple performance indicators since a diverse group of stakeholders with conflicting interests is involved in the internal operating room supply chain.

Abstract: The upcoming fifth generation (5G) wireless communication system is expected to support a broad range of newly emerging applications on top of the regular cellular mobile broadband services. One of the key usage scenarios in the scope of 5G is ultra-reliable and low-latency communications (URLLC). Among the active researchers from both academy and industry, one common view is that URLLC will play an essential role in providing connectivity for the new services and applications from vertical domains, such as factory automation, autonomous driving and so on. The most important key performance indicators (KPIs) related to URLLC are latency, reliability and availability. In this paper, after brief discussion on the design challenges related to URLLC use cases, we present an overview of the available technology components from 3GPP Rel-15 and potential ones from Rel-16. In addition, coordinated multi-cell resource allocation methods are studied. From the system level simulation results in an urban macro environment, it can be observed that effective multi-cell cooperation, more specifically soft combining, can lead to a significant gain in terms of URLLC capacity.

Abstract: Energy sustainability is one of the key questions that drive the debate on cities’ and urban areas development. In parallel, artificial intelligence and cognitive computing have emerged as catalysts in the process aimed at designing and optimizing smart services’ supply and utilization in urban space. The latter are paramount in the domain of energy provision and consumption. This paper offers an insight into pilot systems and prototypes that showcase in which ways artificial intelligence can offer critical support in the process of attaining energy sustainability in smart cities. To this end, this paper examines smart metering and non-intrusive load monitoring (NILM) to make a case for the latter’s value added in context of profiling electric appliances’ electricity consumption. By employing the findings in context of smart cities research, the paper then adds to the debate on energy sustainability in urban space. Existing research tends to be limited by data granularity (not in high frequency) and consideration of about six kinds of appliances. In this paper, a hybrid genetic algorithm support vector machine multiple kernel learning approach (GA-SVM-MKL) is proposed for NILM, with consideration of 20 kinds of appliance. Genetic algorithm helps to solve the multi-objective optimization problem and design the optimal kernel function based on various kernel properties. The performance indicators are sensitivity (Se), specificity (Sp) and overall accuracy (OA) of the classifier. First, the performance evaluation of proposed GA-SVM-MKL achieves Se of 92.1%, Sp of 91.5% and OA of 91.8%. Second, the percentage improvement of performance indicators using proposed method is more than 21% compared with traditional kernel. Third, results reveal that by keeping different modes of electric appliance as identical class label, the performance indicators can increase to about 15%. Forth, tunable modes of GA-SVM-MKL classifier are proposed to further enhance the performance indicators up to 7%. Overall, this paper is a bold and novel contribution to the debate on energy utilization and sustainability in urban spaces as it integrates insights from artificial intelligence, IoT, and big data analytics and queries them in a context defined by energy sustainability in smart cities.

Abstract: In this paper, we present a 4G trace dataset composed of client-side cellular key performance indicators (KPIs) collected from two major Irish mobile operators, across different mobility patterns (static, pedestrian, car, bus and train). The 4G trace dataset contains 135 traces, with an average duration of fifteen minutes per trace, with viewable throughput ranging from 0 to 173 Mbit/s at a granularity of one sample per second. Our traces are generated from a well-known non-rooted Android network monitoring application, G-NetTrack Pro. This tool enables capturing various channel related KPIs, context-related metrics, downlink and uplink throughput, and also cell-related information. To the best of our knowledge, this is the first publicly available dataset that contains throughput, channel and context information for 4G networks.To supplement our real-time 4G production network dataset, we also provide a synthetic dataset generated from a large-scale 4G ns-3 simulation that includes one hundred users randomly scattered across a seven-cell cluster. The purpose of this dataset is to provide additional information (such as competing metrics for users connected to the same cell), thus providing otherwise unavailable information about the eNodeB environment and scheduling principle, to end user. In addition to this dataset, we also provide the code and context information to allow other researchers to generate their own synthetic datasets.

Abstract: Logs are often used for troubleshooting in large-scale software systems. For a cloud-based online system that provides 24/7 service, a huge number of logs could be generated every day. However, these logs are highly imbalanced in general, because most logs indicate normal system operations, and only a small percentage of logs reveal impactful problems. Problems that lead to the decline of system KPIs (Key Performance Indicators) are impactful and should be fixed by engineers with a high priority. Furthermore, there are various types of system problems, which are hard to be distinguished manually. In this paper, we propose Log3C, a novel clustering-based approach to promptly and precisely identify impactful system problems, by utilizing both log sequences (a sequence of log events) and system KPIs. More specifically, we design a novel cascading clustering algorithm, which can greatly save the clustering time while keeping high accuracy by iteratively sampling, clustering, and matching log sequences. We then identify the impactful problems by correlating the clusters of log sequences with system KPIs. Log3C is evaluated on real-world log data collected from an online service system at Microsoft, and the results confirm its effectiveness and efficiency. Furthermore, our approach has been successfully applied in industrial practice.

Abstract: Consumer recommendations of products and services are important performance indicators for organizations to gain feedback on their offerings. Furthermore, they are important for prospective customers to learn from prior consumer experiences. In this study, we focus on user-generated content, in particular online reviews, to investigate which service aspects are evaluated by consumers and how these factors explain a consumer's recommendation. Further, we investigate how recommendations can be predicted automatically based on such user-driven responses. We disentangle the recommendation decision by performing explanatory and predictive analyses focusing on a sample of airline reviews. We identify core and augmented service aspects expressed in the online review. We then show that service aspect-specific sentiment indicators drive the decision to recommend an airline and that these factors can be incorporated in a predictive model using data mining techniques. We also find that the business model of an airline being reviewed, whether low cost or full service, is also an applicable consideration. Our results are highly relevant for practitioners to analyze and act on consumer feedback in a prompt manner, along with the ability of gaining a deeper understanding of the service from multiple aspects. Also, potential travelers can benefit from this approach by getting an aggregated view on service quality.

Abstract: Sustainability has become a necessity, partly due to the threats created by traditional manufacturing practices, and due to regulations imposed by stakeholders. Performance evaluation is an important component of sustainability initiatives in manufacturing organizations. This study proposes a sustainability evaluation method for manufacturing SMEs using integrated fuzzy analytical hierarchal process (FAHP) and fuzzy inference system (FIS) approach. The performance indicators are identified from literature considering the characteristics of SMEs. Balanced scorecard framework is used to categorize the indicators among its four aspects. The linguistic variables are used to collect the opinions of decision makers about the performance ratings and importance of the aspects and corresponding indicators. The FAHP method is applied to determine the relative weights of measures and indicators. The performance ratings of the organization with respect to indicators and relative weights of indicators are combined to obtain the weighted performance ratings. The weighted performance ratings are considered as inputs to FIS. The hierarchal FIS is applied to derive the overall sustainability performance. Using a case study of manufacturing SME, the sustainability score of the organization was elicited in accordance with this procedure. Consequently, a sensitivity analysis of the proposed method reveals the most important basic indicators affecting overall sustainability, identifying areas which decision makers should place special attention. This method can also assist managers of larger enterprises to assess the effectiveness of their sustainability strategies, especially when dealing with suppliers from the SMEs.

Abstract: Reverse logistics initiatives with social commerce not only provide opportunities for firms to create new sources of revenue but also demonstrate their corporate social responsibility via social, green, and environmental activities. Thus, a growing number of companies are attempting to streamline their social commerce platforms to effectively handle reverse logistics. The purpose of this study is to identify the criteria that should be used in designing and evaluating social commerce based reverse logistics processes by firms. We tested the effectiveness of the identified criteria by using them to evaluate the reverse logistics practices of three major global firms that use social commerce platforms. First, we identified the criteria from a thorough review of the literature. Then, we invited five experts to provide (linguistic) ratings of these firms on the selected criteria, using a fuzzy Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) technique with FLINTSTONES (a software tool) to generate aggregate scores for the assessment and evaluation of reverse logistics practices in social commerce platforms. Sensitivity analysis was also provided to monitor the robustness of the approach. The results of the study identified that four dominant criteria (reverse logistics performance indicators) in the social commerce platform: Customer relationship, Usage risk, Reviews, and Quality control.

Abstract: Additive manufacturing (AM) is a very promising technology; however, there are a number of open issues related to the different AM processes. The literature on modelling the existing AM processes is reviewed and classified. A categorization of the different AM processes in process groups, according to the process mechanism, has been conducted and the most important issues are stated. Suggestions are made as to which approach is more appropriate according to the key performance indicator desired to be modelled and a discussion is included as to the way that future modelling work can better contribute to improving today's AM process understanding.

Abstract: Vehicle-to-everything (V2X) communication enables vehicles, roadside vulnerable users, and infrastructure facilities to communicate in an ad-hoc fashion. Cellular V2X (C-V2X), which was introduced in the 3rd generation partnership project (3GPP) release 14 standard, has recently received significant attention due to its perceived ability to address the scalability and reliability requirements of vehicular safety applications. In this paper, we provide a comprehensive study of the resource allocation of the C-V2X multiple access mechanism for high-density vehicular networks, as it can strongly impact the key performance indicators such as latency and packet delivery rate. Phenomena that can affect the communication performance are investigated and a detailed analysis of the cases that can cause possible performance degradation or system limitations, is provided. The results indicate that a unified system configuration may be necessary for all vehicles, as it is mandated for IEEE 802.11p, in order to obtain the optimum performance. In the end, we show the inter-dependence of different parameters on the resource allocation procedure with the aid of our high fidelity simulator.

Abstract: The continuous digitization requires organizations to improve the automation of their business processes. Among others, this has lead to an increased interest in Robotic Process Automation (RPA). RPA solutions emerge in the form of software that automatically executes repetitive and routine tasks. While the benefits of RPA on cost savings and other relevant performance indicators have been demonstrated in different contexts, one of the key challenges for RPA endeavors is to effectively identify processes and tasks that are suitable for automation. Textual process descriptions, such as work instructions, provide rich and important insights about this matter. However, organizations often maintain hundreds or even thousands of them, which makes a manual analysis unfeasible for larger organizations. Recognizing the large manual effort required to determine the current degree of automation in an organization’s business processes, we use this paper to propose an approach that is able to automatically do so. More specifically, we leverage supervised machine learning to automatically identify whether a task described in a textual process description is manual, an interaction of a human with an information system or automated. An evaluation with a set of 424 activities from a total of 47 textual process descriptions demonstrates that our approach produces satisfactory results.

Abstract: Accurate and reliable prediction of project performance is critical to the success of construction projects and companies alike. Such prediction assists in obtaining early warnings against potential problems. Existing project performance forecasting models are fragmented, especially regarding the consideration of interdependency between multidimensional performance indicators. To address these limitations, a system dynamics (SD) model was developed to simulate the complexities that exist among interdependent variables and forecast their dynamics over time. The proposed model integrates eight construction project performance indices, which have been identified through literature review and interviews with domain experts. Performance dimensions include cost, schedule, quality, profitability, safety, environment, team satisfaction, and client satisfaction. This model focuses on the construction phase of projects under unit price contracts and is intended for use by contractors. The model was tested on a road construction project to assess its practicality and accuracy. Results demonstrate general agreement between actual and forecasted performance indices. The model was also used to simulate four possible intervention scenarios by the project manager. Results of various scenarios show overall agreement with expected impacts of the interventions. The research advances the state of practice and knowledge of project performance forecasting through the creation of a more holistic and interdependent model of project performance metrics.

Abstract: Objectives Internationally, there has been considerable debate about the role of data in supporting quality improvement in health care. Our objective was to understand how, why and in what circumstances the feedback of aggregated patient-reported outcome measures data improved patient care. Methods We conducted a realist synthesis. We identified three main programme theories underlying the use of patient-reported outcome measures as a quality improvement strategy and expressed them as nine 'if then' propositions. We identified international evidence to test these propositions through searches of electronic databases and citation tracking, and supplemented our synthesis with evidence from similar forms of performance data. We synthesized this evidence through comparing the mechanisms and impact of patient-reported outcome measures and other performance data on quality improvement in different contexts. Results Three programme theories were identified: supporting patient choice, improving accountability and enabling providers to compare their performance with others. Relevant contextual factors were extent of public disclosure, use of financial incentives, perceived credibility of the data and the practicality of the results. Available evidence suggests that patients or their agents rarely use any published performance data when selecting a provider. The perceived motivation behind public reporting is an important determinant of how providers respond. When clinicians perceived that performance indicators were not credible but were incentivized to collect them, gaming or manipulation of data occurred. Outcome data do not provide information on the cause of poor care: providers needed to integrate and interpret patient-reported outcome measures and other outcome data in the context of other data. Lack of timeliness of performance data constrains their impact. Conclusions Although there is only limited research evidence to support some widely held theories of how aggregated patient-reported outcome measures data stimulate quality improvement, several lessons emerge from interventions sharing the same programme theories to help guide the increasing use of these measures.

Abstract: Purpose The purpose of this paper is to propose an integrated performance measurement framework to measure the effect of lean implementation throughout all functions of an organization. Design/methodology/approach The paper identifies the seven categories representing all organizational functions. These categories have been divided into 26 performance dimensions and key performance indicators (KPIs) for each performance dimension have been identified to measure lean performance. The interrelationship of each category with lean principles and/or lean wastes has been identified. KPIs are developed on the basis of identified criteria, frequency analysis of existing literature, and discussion with industry professionals. Finally, an integrated performance measurement framework is proposed. Findings The proposed framework evaluates the organization under seven categories – manufacturing process, new product development (NPD), human resource management, finance, administration, customer management, and supplier management. In total, 26 dimensions and 119 key performance indicators have been identified under the seven categories. Research limitations/implications The proposed framework is a conceptual framework and it is to be tested by empirical and cross-sectional studies. Originality/value The main novelty of the research is that the leanness of the organization has been measured throughout the supply chain of the organization in an integrated way. The various areas of measurement are manufacturing process, NPD, finance, administration, customer management, and supplier management. Further, the proposed KPIs are also categorized as qualitative or quantitative, strategic or operational, social or technical, financial or non-financial, leading or lagging, static or dynamic. This paper contributes to the body of knowledge in performance measurement.