Multi-stage Service Research Articles

Design of Patient Visit Itineraries in Tandem Systems

Problem definition: Multistage service is common in healthcare. One widely adopted approach to manage patient visits in multistage service is to provide patients with visit itineraries that specify personalized appointment time for each patient at each service stage. We study how to design such visit itineraries. Methodology/results: We develop the first optimization modeling framework to provide each patient with a personalized visit itinerary in a tandem (healthcare) service system. Due to interdependencies among stages, our model loses those elegant properties (e.g., L-convexity and submodularity) often utilized to solve the classic single-stage models. To address these challenges, we develop two original reformulations. One is directly amenable to off-the-shelf optimization software, and the other is a concave minimization problem over a polyhedron shown to have neat structural properties, based on which we develop efficient solution algorithms. In addition to these exact solution approaches, we propose an approximation approach with a provable optimality bound and numerically validated performance to serve as an easy-to-implement heuristic. A case study populated by data from the Dana-Farber Cancer Institute shows that our approach makes a remarkable 28% cost reduction over practice on average. Managerial implications: Common approaches used in practice are based on simple adjustments to schedules generated by single-stage models, often assuming deterministic service times. Whereas such approaches are intuitive and take advantage of existing knowledge of single-stage models, they can lead to significant loss of operational efficiency in managing multistage services. A well-designed patient visit itinerary that carefully addresses the interdependencies among stages can significantly improve patient experience and provider utilization. History: This paper was selected for Fast Track in the M&SOM Journal from the 2022 MSOM Healthcare SIG Conference. Funding: The work of the last two authors was supported in part by the National Natural Science Foundation of China [Grants 71931008 and 72001220]. Supplemental Material: The online appendix is available at https://doi.org/10.1287/msom.2023.0134 .

Multi-windowing Technique for 5G and Beyond

Recently, the development of mobile technology has reached a point that requires a high data rate with high waveform aspects. So, traditional ways are no further tolerable. With the dawn of the fifth generation (5G), researchers have been investigating its waveforms and experimenting the methods to improve it. This research examines universal filter multi-carrier (UFMC) for a novel strategy by employing multi-windowing UFMC-based (MW-UFMC) to boost the power spectral density (PSD). This was accomplished by lowering out-of-band emission (OOBE), bit error rate (BER), and adjacent channel leakage ratio (ACLR). The 256quadrature amplitude modulation (256QAM) mapping modulation method was employed in this work. 256QAM is the maximum modulation level that is standard for the third-generation partnership project (3GPP). The proposed MW-UFMC is compared with the conventional UFMC and cyclic prefix orthogonal frequency division multiplexing (CP-OFDM). The proposed multi-windowing technique presents a multistage service that offer a diversity of options to the network operator, which exhibits several advantages over existing methods. These include a lower BER, improved PSD, enhanced spectral efficiency (SE), reduced OOBE that increases capacity for accommodating more bits, decreased ACLR, and lower latency.

Sequencing and scheduling appointments for multi-stage service systems with stochastic service durations and no-shows

ABSTRACT In this work, we consider a joint sequencing and scheduling appointments problem with stochastic service times and no-shows in multi-stage service systems. The objective is to minimise the total expected weighted costs of customers' waiting times and service providers' idle times over multiple stages. For the problem, we first formulate it as a stochastic program and exploit the sample average approximation approach to reformulate it as a mixed-integer program in further. Then we transform the stochastic program into a two-stage optimisation problem and develop a standard Benders decomposition algorithm. To overcome the long running time of Benders decomposition, we simplify the master problem in the algorithm and propose a Benders decomposition-based algorithm to find a near-optimal solution. Finally, we conduct a series of numerical experiments to illustrate the efficiency of our proposed algorithm, examine the impact of the number of stages, stochastic service times and no-shows on the optimal job allowances and performance indicators (i.e. waiting times and idle times) and investigate two easy-to-implement sequence rules. The computational results show that both our proposed Benders decomposition-based algorithm and easy-to-implement sequence rules perform well.

MuSC: A multi-stage service chains embedding approach

Network function virtualization is an emerging concept that is attracting increasing attention in the industry because it offers high levels of agility and flexibility to the network allowing easier software updates, resource adjustment and scalable changes in the configuration of the services. Network functions traditionally delivered on hardware and purpose-built platforms in legacy networks can now be provided through shared virtual resources called VNFs (Virtual network functions) hosted over shared physical network infrastructures. Using network functions as absolute software components would certainly improve the deployment process and the management of the VNF life cycle. Since VNFs are hardware-independent, there is no need to check whether or not a network function is compatible with the rest of the network's physical parts; this also helps reduce the maintenance costs, which can increase drastically in the case of major upgrades to the network infrastructure. Along with all these benefits, shifting to virtual-based networks comes with a significant number of challenges, especially in terms of placing such virtual components, ensuring their interoperability and maintaining the quality of service at a level that is at least as good as what is offered by hardware based architectures. In this paper, we propose a cluster-based placement and chaining solution. The overall proposed approach consists of: 1) formulating an Integer Linear Programming (ILP) model aimed at finding an optimal tradeoff between multiple objective functions that might be sometimes conflicting (e.g hardware resource and energy consumption minimization, transmission delays, bandwidth usage, etc…), 2) classifying the substrate network into a set of on-demand clusters that are efficient for a predefined set of metrics, and 3) using meta-heuristic-based algorithms to find near-optimal solutions for the formulated ILP.

Physician Staffing in Emergency Rooms (ERs): Opening the Black-box of ER Care via a Multi-Class Multi-Stage Network

Problem Definition and Relevance: Characterized by time-varying arrivals, multi-stage service, and multi-class patient population, emergency rooms (ERs) are complex healthcare delivery systems, where optimizing the staffing levels of physicians is a challenge. In collaboration with Mayo Clinic, we study a staffing and an associated routing problem for ER physicians and propose a new staffing rule to meet tail probability of delay (TPoD) type service targets. Methodology: We capture the time-varying patient flow in the ER with a multi-class multi-stage queuing network, describing the ER care as sequences of treatment queues, where physicians serve, and groups of diagnostic medical processes. Treatment stations in ER are busy-server queues, where waiting before service is common, but experience negligible abandonments. Motivated by these queues, we propose a new staffing algorithm, translating the offered load into staffing decisions for efficiency-driven queues with TPoD targets and perfectly patient customers. Results: We analytically show the asymptotic effectiveness of our staffing rule on stabilizing TPoD for M/M/s queues, operating in efficiency-driven mode, and numerically demonstrate its robustness and optimality in various time-varying ER settings via realistic and data-driven simulation experiments. We further show that as the service complexity of an ER increases, hybrid routing rules, using pre-determined (static) priorities and (dynamic) current system state jointly, become necessary to meet TPoD targets. Managerial Implications: Instead of treating the entire patient length of stay as a black-box service model, our multi-stage network model provides more managerial control over the internal components of ER service and is easily implementable in practice.

Optimal control policies in service systems with limited information on the downstream stage

AbstractExisting models in multistage service systems assume full information on the state of downstream stages. In this paper, we investigate how much the lack of such information impacts jobs' waiting time in a two‐stage system with two types of jobs at the first stage. The goal is to find the optimal control policy for the server at the first stage to switch between type‐1 and type‐2 jobs, while minimizing the long‐run average number of jobs in the system. We identify control policies and corresponding conditions under which having no or partial information, the system can still capture the most benefit of having full information.

Control charts for monitoring multi-stage service processes with optimal queue performance

This paper develops control charts for monitoring the service times at the different stages of a congested system that provides a multi-stage service to its customers. A simulation-based optimization procedure is used to simulate the whole service process and determine the chart parameters that minimize the average time a customer spends in the system. Statistical constraints are also considered to enhance the statistical properties of the control chart. The proposed socio-statistical method may be of interest for a variety of applications. The method is applied to a fast food restaurant with two waiting stages.

Heuristic Algorithm for Forming the Service Schedule for the Multistage Systems with a Generalized Resource Criterion

The paper suggests one of the approaches to solving the problem of planning multi-stage service systems, the distinguishing feature of which is the availability of time constraints and a resource criterion. At the heart of the approach lie heuristics, which make it possible to obtain a suboptimal solution in an acceptable time. Based on these heuristics, the process of making decisions about assigning jobs to a given time interval, determining the optimal time interval for the return, and estimating the remaining service time are carried out.

Early Task Initiation and Other Load-Adaptive Mechanisms in the Emergency Department

We study a multistage service process that adapts to system occupancy level. Using operational data from more than 140,000 patient visits to a hospital emergency department, we show that the system-level performance of the emergency department is an aggregation of several simultaneous server-level workload response mechanisms. We identify early task initiation as a between-stage adaptive response mechanism that occurs when an upstream stage initiates tasks that are normally handled by a downstream stage. We show that having some diagnostic tests ordered during the triage process reduces treatment time by 20 minutes, on average. However, ordering too many tests at triage can lead to an increase in the total number of tests performed on the patient. We also demonstrate the presence of other response mechanisms such as queuing delays for tasks such as medication delivery, and rushing as nurses spend less time with their patients when the queue length is high. This paper was accepted by Serguei Netessine, operations management.

A novel service discovery algorithm via service repository centre in internet of things

In recent years, the internet of things (IoT) has attracted more and more attention, and IoT services have increased rapidly. The service discovery which has important implications for the service integration of IoT services has become more difficult in large-scale registration. In traditional matching method, in order to get a better match results, all the matching parameters for services had to be calculated together, thus it would waste a lot of computing resources and computing time. This paper propose a centralise four-layer model for service discovery via multi-stage semantic service matching to improve the performance of service discovery, namely interactive interface layer, parsing annotation layer, service matching layer and data semantic layer, so as to meet the requirement of service discovery in large-scale IoT services. We also propose a hybrid service matching degree measurement by synthetically calculating the concept logic and semantic similarity for each layer singly. Experimental analysis confirms the effectiveness of the proposed mechanism in improving the precision and recall.

Transient Analysis of Markovian Queue Model with Multi Stage Service

The present investigation deals with the transient analysis of Markovian queue model with multi stage service. Here, current analysis deals with M/M/1 queueing model with finite capacity. The arrival of the customer in the system follows Poisson process and the customers are served following exponential distribution. The brokendown server is immediately send for repair process so as to become as good as earlier. Transient analysis of finite Markovian queueing model has been studied using Runge-Kutta method. Various performance measures like queue length of the system, waiting time, throughput, service state probabilities have also been obtained.

Waiting in multi-stage services: an exploration across service industries

This field study explores the relative importance of waiting time perception with waits before, during and following the actual service delivery in three multi–stage services. Data were collected in various service settings, i.e., in restaurants, assessment and dental services in the Netherlands. Structural equation modelling is used to analyse the data. Results imply that in restaurant services, the service–entry wait has the largest influence on service evaluation, in the assessment services the service–exit wait is predominant and in the dental services, the in–service wait is the most important wait. Researchers and practitioners are suggested to apply a stage approach when considering waiting time satisfaction in their work. This study is one of the first to empirically explore the effect of service waits in multiple succeeding services stages outside restaurant services.

A framework for service quality improvement in multi-stage processes

The purpose of this research is to develop an integrative framework for identifying and prioritising process improvement opportunities in a multi-stage service operation with the objective of improving quality perception. The proposed framework identifies important perceptual elements in the process and links the abstract nature of service quality perception with the operational measures (e.g., delivery time) of actual process performance. The proposed framework looks beyond customer SQ perception scores and analyses the actual performance of the process alongside customer perception in a multi-stage service process. The framework recommends that the customer perception of service quality and process performance are to be measured and analysed simultaneously for any process improvements. The applicability of the framework is verified by using a real-life case on customer service request and the service delivery process. Ordinal logistic regression is used to determine the process improvement priorities. The case study findings provide evidence and critical insight into the linkage of customer perception of service quality with operational measures of actual process performance.

Genetic Algorithm for Inventory Levels and Routing Structure Optimization in Two Stage Supply Chain

Several analytical models have been developed to solve the integrated production distribution problems in Supply Chain Management (SCM). In certain multi-stage service supply chain like blood banks, the term ‘production’ is referred as collection. It is often crucial to consider the inventory and distribution costs for successful decision making in multi-stage service supply chain. In this paper, the authors have explored this problem by considering a Two - Stage Collection - Distribution (TSCD) Model for blood collection and distribution that faces a deterministic stream of external demands for blood product. A finite supply and collection of blood at stage one Central Blood Bank (CBB) has been assumed. Blood is collected at stage one CBB and distributed to stage two Regional Blood Bank (RBB), where the storage capacity of the RBB is limited. Packaging is completed at stage two (that is, value is added to each item, but no new items are created), and the packed blood bags are stored which is used to meet the final demand of customer zone. During each period, the optimal collection rate at CBB, distribution rate between CBB and RBB and routing structure from the CBB to RBB and then to customer zone, must be determined. This TSCD model with capacity constraints at both stages is optimized using Genetic Algorithms (GA) and compared with the standard operations research software LINDO for small problems.

A six sigma approach to boost up time domain reliability in multi-stage services

Reliability function, R(t), is defined as the probability that the system will not fail during the stated period of time under a given operating conditions. In production process, attaining reliability as a function of time is a common task but in multi-stage services, it is not straightforward due to intricacy of gathering data. Through the present work, we have addressed a systematic method for measuring multi-stage service reliability function using failure rate analysis beside a systematic six sigma approach to improve total system reliability. The proposed methodology furthermore could help practitioners to integrate six sigma metrics with the reliability function and the way to measure existing reliability situation based on sigma level. The proposed method is applied in a given multi-stage central repairing shop for an automobile producer which promotes 86% in total system reliability beside 58% improvement in the sigma level. Key words: Service reliability, six sigma, failure rate, multi-stage processes, sigma level.

Dynamic Voltage Scaling in Multitier Web Servers with End-to-End Delay Control

The energy and cooling costs of Web server farms are among their main financial expenditures. This paper explores the benefits of dynamic voltage scaling (DVS) for power management in server farms. Unlike previous work, which addressed DVS on individual servers and on load-balanced server replicas, this paper addresses DVS in multistage service pipelines. Contemporary Web server installations typically adopt a three-tier architecture in which the first tier presents a Web interface, the second executes scripts that implement business logic, and the third serves database accesses. From a user's perspective, only the end-to-end response across the entire pipeline is relevant. This paper presents a rigorous optimization methodology and an algorithm for minimizing the total energy expenditure of the multistage pipeline subject to soft end-to-end response-time constraints. A distributed power management service is designed and evaluated on a real three-tier server prototype for coordinating DVS settings in a way that minimizes global energy consumption while meeting end-to-end delay constraints. The service is shown to consume as much as 30 percent less energy compared to the default (Linux) energy saving policy

Service Competition with General Queueing Facilities

In many service industries, companies compete with each other on the basis of the waiting time their customers' experience, along with the price they charge for their service. A firm's waiting time standard may either be defined in terms of the expected value or a given, for instance 95%, percentile of the steady state waiting time distribution. We investigate how a service industry's competitive behavior depends on the characteristics of the service providers' queueing systems. We provide a unifying approach to investigate various standard single stage systems covering the spectrum from M/M/1 to general G/GI/s systems, along with open Jackson networks to represent multi-stage service systems. Assuming that the capacity cost is proportional with the service rates we refer to its dependence on (i) the firm's demand rate and (ii) the waiting time standard as the capacity cost function. We show that across the above road spectrum of queueing models, the capacity cost function belongs to a specific four parameter class of function, either exactly or as a close approximation. We then characterize how this capacity cost function impacts on the equilibrium behavior in the industry. We give separate treatments to the case where the firms compete in terms of (i) prices (only) (ii) their service level or waiting time standard (only), (iii) simultaneously in terms of both prices and service levels. The firms' demand rates are given by a general system of equations of the prices and waiting time standards in the industry.

Scheduling multi-stage parallel-processor services to minimize average response time

The problem of scheduling on a multi-stage parallel-processor architecture in computer centres is addressed with the objective of minimizing average completion time of a set of requests. The problem is modelled as a flexible flowshop problem for which few heuristics exist in the flowshop scheduling literature. A new three-phase heuristic is proposed in this paper. An extensive computational experiment has been conducted to compare the performance of the existing heuristics and the proposed heuristic. The results indicate that the proposed heuristic significantly outperforms the existing ones. More specifically, the overall average error of the best existing heuristic is about five times that of the proposed heuristic while the overall average CPU time of the proposed heuristic is about half of the best existing one. More importantly, as the number of requests increases, the CPU time of the proposed heuristic decreases considerably (compared to the best existing heuristic) while the ratio of the error (of the best existing to the proposed heuristic) of about five times remains almost the same.

Measuring performance in multi-stage service operations: An application of cause selecting control charts

Many multistage service operations exhibit the cascade property, where performance at one stage is statistically correlated with performance at the preceding stage. Prior research on multistage services has analyzed each process stage independently or in an additive manner. Increased emphasis on Six Sigma initiatives in services has rekindled interest in the use of control charts to monitor and control service processes. This study examines the cause selecting control chart as a methodology to monitor and identify potential problem areas in an actual cascade service process and compares the diagnostic capability of the cause selecting chart to that of a traditional Shewhart chart. A grocery store whose parent company was implementing efficient consumer response (ECR) serves as the research context. This study models the grocery store as a two-stage cascade process and uses operating data from the store to construct a cause selecting chart and a traditional Shewhart chart for the front-end operation. Analysis of the two charts reveals that the cause selecting chart outperforms the traditional control chart as tool for signaling unusual variation in performance at the front-end stage. The analysis demonstrates that service managers can receive misleading or erroneous information from traditional control charts if the service process being monitored is a cascade process.

About the cumulative idle time in multiphase queues

The paper is designated to the analysis of queueing systems, arising in the network theory and communications theory (called multiphase queueing systems, tandem queues or series of queueing systems). Also we note that multiphase queueing systems can be useful for modelling practical multi-stage service systems in a variety of disciplines, especially on manufacturing (assembly lines), computer networking (packet switch structures), and in telecommunications (e.g. cellular mobile networks), etc. This research presents heavy traffic limit theorems for the cumulative idle time in multiphase queues. In this work, functional limit theorems are proved for the values of important probability characteristics of the queueing system (a cumulative idle time of a customer).