Inneke Van Nieuwenhuyse

KU Leuven | ku leuven · Research Centre for Management Informatics (LIRIS)

We present a Markov model to analyze the queueing behavior of the nonstationary G(t)/G(t)/s(t)+G(t) queue. We assume an exhaustive service discipline (where servers complete their current service before leaving) and use acyclic phase-type distributions to approximate the general interarrival, service, and abandonment time distributions. The time-va...

This paper presents a newsvendor approach to determine optimal order-up-to levels in a two-item inventory system with one-way substitution, assuming that both items are managed according to a periodic base stock order policy. The objective is to minimize the expected total cost per period, which consists of expected purchasing costs, expected inven...

This article uses a sequentialized experimental design to select simulation input combinations for global optimization, based on Kriging (also called Gaussian process or spatial correlation modeling); this Kriging is used to analyze the input/output data of the simulation model (computer code). This design and analysis adapt the classic “expected i...

This article presents a novel heuristic for constrained optimization of computationally expensive random simulation models. One output is selected as objective to be minimized, while other outputs must satisfy given threshold values. Moreover, the simulation inputs must be integer and satisfy linear or nonlinear constraints. The heuristic combines...

Hyperparameter optimization (HPO) is a necessary step to ensure the best possible performance of Machine Learning (ML) algorithms. Several methods have been developed to perform HPO; most of these are focused on optimizing one performance measure (usually an error-based measure), and the literature on such single-objective HPO problems is vast. Rec...

The performance of any Machine Learning algorithm is impacted by the choice of its hyperparameters. As training and evaluating a ML algorithm is usually expensive, the hyperparameter optimization (HPO) method needs to be computationally efficient to be useful in practice. Most of the existing approaches on multi-objective HPO use evolutionary strat...

In this research, Artificial Intelligence (AI) was used to support the optimization of six bonding process parameters for maximal joint strength and minimal production costs. Two industrial bonding processes were investigated, one from electronic potting and another from the manufacturing industry. The focus was on optimizing the plasma treatment o...

The performance of any Machine Learning (ML) algorithm is impacted by the choice of its hyperparameters. As training and evaluating a ML algorithm is usually expensive, the hyperparameter optimization (HPO) method needs to be computationally efficient to be useful in practice. Most of the existing approaches on multi-objective HPO use evolutionary...

We consider multiobjective simulation optimization problems, where several conflicting objectives are optimized simultaneously, and can only be observed via stochastic simulation. The goal is to find or approximate a (discrete) set of Pareto-optimal solutions that reveal the essential trade-offs between the objectives, where optimality means that n...

Multi-objective optimization of complex engineering systems is a challenging problem. The design goals can exhibit dynamic and nonlinear behaviour with respect to the system’s parameters. Additionally, modern engineering is driven by simulation-based design which can be computationally expensive due to the complexity of the system under study. Baye...

Product recalls are highly disruptive for many firms. Understanding the drivers of such recalls is paramount to helping firms effectively reduce product recall risk. While prior studies have investigated the drivers of product recalls in developed markets, little is known about the factors that drive product recalls in emerging markets. Using data...

Automotive companies are increasingly looking for ways to make their products lighter, using novel materials and novel bonding processes to join these materials together. Finding the optimal process parameters for such adhesive bonding process is challenging. In this research, we successfully applied Bayesian optimization using Gaussian Process Reg...

Hyperparameter optimization (HPO) is a necessary step to ensure the best possible performance of Machine Learning (ML) algorithms. Several methods have been developed to perform HPO; most of these are focused on optimizing one performance measure (usually an error-based measure), and the literature on such single-objective HPO problems is vast. Rec...

To cost effectively offer distinctive products to the market, many companies resort to product platforms, defined as a common base from which a set of products are derived. Yet, the interaction between both types of decisions (product development versus product portfolio design) is poorly understood. Indeed, operational aspects (such as platform pr...

Automotive companies are increasingly looking for ways to make their products lighter, using novel materials and novel bonding processes to join these materials together. Finding the optimal process parameters for such adhesive bonding process is challenging. In this research, we successfully applied Bayesian optimization using Gaussian Process Reg...

The use of kriging metamodels in simulation optimization has become increasingly popular during recent years. The majority of the algorithms so far uses the ordinary (deterministic) kriging approach for constructing the metamodel, assuming that solutions have been sampled with infinite precision. This is a major issue when the simulation problem is...

This article surveys the most relevant kriging-based infill algorithms for multiobjective simulation optimization. These algorithms perform a sequential search of so-called infill points, used to update the kriging metamodel at each iteration. An infill criterion helps to balance local exploitation and global exploration during this search by using...

Motivation Patients with hematological malignancies are susceptible to life-threatening infections after chemotherapy. The current study aimed to evaluate whether management of such patients in dedicated inpatient and emergency wards could provide superior infection prevention and outcome. Methods We have developed an approach allowing to retrieve...

Distribution of the WBC counts among the evaluated patients based on different factors (disease, protocol length, number of treatment cycles, hospitalization, infection occurrence, and patient age). (PDF)

Overview of cytotoxic and non-cytotoxic anti-neoplastic drugs applied in the study population. (PDF)

Mortality after infection: Model selection and sensitivity analysis. Table A. Comparison of models for mortality after infection analysis: Probit, logit, and cloglog Table B. The location as an explanatory variable in the model for mortality after infection analysis Table C. Cycle number in the model for mortality after infection analysis Table D....

List of antibiotics commonly used for prophylaxis or active infection therapy in the study population. (PDF)

Infection after protocol completion: Model selection and sensitivity analysis. Table A. Comparison of models for infection after protocol completion analysis: Probit, logit and cloglog Table B. The location as an explanatory variable in the model for infection after protocol completion analysis Table C. Cycle number in the model for infection after...

Hospitalization delay by entry gate to the hospital (ED vs. HOutC) and the hour of admission. (PDF)

Emergency departments (EDs) are one of the main entry points of a hospital, offering non-stop healthcare services to patients with various needs. ED crowding is considered a major international problem. To cope with this problem, operations research techniques have been widely applied to analyse and optimise ED operations. In this regard, two essen...

Bayesian network is a kind of uncertainty knowledge expression and reasoning tool, and it is an effective means to solve problems in related fields such as information retrieval. Considering the characteristics of e-commerce supply chain supply information and Bayesian network, a cognitive big data analysis method for intelligent information system...

Supply chain management is a kind of behavior for enterprises to create core competitiveness in a complex competitive environment, especially for supply chain finance, which is a typical complex system with dynamic, open, and emergent non-linear characteristics in structure, environment, and behavior. For fractional order non-linear multi-agent sys...

In this article, we consider the impact of finite production capacity on the optimal quality and pricing decisions of a make-to-stock manufacturer. Products are differentiated along a quality index; depending on the price and quality levels of the products offered, customers decide to either buy a given product, or not to buy at all. We show that,...

This study seeks insights into the impact of inpatient boarding on emergency department (ED) congestion and capacity. To do so, we model the ED as a semi-open queueing network (SOQN) with limited resources (physicians and beds) and discontinuous patient service. We present a Markov-modulated fluid queue approach to efficiently calculate service lev...

As an underlying support technology, blockchain is a shared ledger system and a computational paradigm, which is decentralized, and it is highly compatible with the distributed economic system. The distributed scheduling model of agricultural business resources based on the public service platform is a comprehensive solution to the current situatio...

In recent years, the problems of low degree of industrialization of agriculture, weak informatization ability and food safety have become increasingly serious. This article combines the detection of agricultural products supply chain and RFID technology and applies it to the testing of agricultural products supply chain. In this study, the agricult...

In this article we investigate the unconstrained optimization (minimization) of the performance of a system that is modeled through a discrete-event simulation. In recent years, several algorithms have been proposed which extend the traditional Kriging-based simulation optimization algorithms (assuming deterministic outputs) to problems with noise....

Over the past decades, the Vehicle Routing Problem (VRP) and its variants have grown ever more popular in the academic literature. Yet, the problem characteristics and assumptions vary widely and few literature reviews have made an effort to classify the existing articles accordingly. In this article, we present a taxonomic review of the VRP litera...

In this article, we present a decision support system (DSS) for improving patient flow in emergency departments (EDs). The core of the system is a discrete-event simulation (DES) model that aims to support capacity planning in the ED, in view of controlling patients' length of stay (LOS). Conceptually, it regards the patient LOS as the result of di...

Simulation optimization is increasingly popular for solving complicated and mathematically intractable business problems. Focusing on academic articles published between 1998 and 2013, the present survey aims to unveil the extent to which simulation optimization has been used for solving practical inventory problems (as opposed to small, theoretica...

This chapter discusses the challenges of workforce scheduling in systems with time-varying arrival rates, with customer impatience (abandonments), general service and abandonment distributions and an exhaustive end-of-shift policy (servers may work overtime at the end of their shift to complete the service process of a customer). We explore the opp...

In recent years, several algorithms have been proposed which extend the traditional Kriging-based simulation optimization methods (assuming deterministic outputs) to problems with noise. Our objective in this paper is to compare the relative performance of a number of these algorithms on a set of well-known analytical test functions, assuming diffe...

In this poster, we evaluate the effectiveness of four kriging-based approaches for simulation optimization with homogeneous noise: Augmented Expected Improvement (AEI), Approximate Knowledge Gradient (AKG), the Two-stage sequential optimization method (TSSO), and the well-known Efficient Global Optimization (EGO) algorithm. We test the performance...

This article presents a straightforward production/inventory model that can capture the trade-offs among average inventory, production capacity and customer service levels in a semi-process industry setting. The model is based on well-known approximations from queuing literature, and it supports midterm planning procedures at SEPPIC, a large specia...

Simulation optimization is increasingly popular for solving complicated and mathematically intractable real-world business problems. This article reviews recent applications of simulation optimization in inventory management. Following a brief discussion of different simulation optimization techniques, this article categorizes recent contributions...

The emergency department (ED) is an interesting field for operations research (OR) and operations management (OM) researchers. Having timevarying arrivals and heterogeneous patients that need to be treated in consecutive processing steps by several doctors, nurses and other employees, it is a complex environment to control. The difficulty to contro...

This article provides details on the modeling and validation of a discrete-event simulation study carried out at the emergency department (ED) of a large regional hospital in Belgium. The ED has 21 beds, and a volume of about 30,000 patients per year of which approximately 33% need to be admitted to the hospital. Like many other hospital EDs all ov...

In many service systems, the arrival pattern is not constant throughout the day. This raises the question how staffing decisions should be adapted in view of controlling customer's waiting times. Assuming a single-stage queueing system with general abandonment and service times and time-varying demand for service, we suggest a method inspired by th...

We present a Markov model to approximate the queueing behavior at the G(t)/G(t)/s(t)+G(t) queue with exhaustive discipline and abandonments. The performance measures of interest are: (1) the average number of customers in queue, (2) the variance of the number of customers in queue, (3) the average number of abandonments and (4) the virtual waiting...

Many shift scheduling algorithms presume that the staffing levels, required to ensure a target customer service, are known in advance. Determining these staffing requirements is often not straightforward, particularly in systems where the arrival rate fluctuates over the day. We present a branch-and-bound approach to estimate optimal shift schedule...

In an emergency department (ED), the demand for service is not constant over time. This cannot be accounted for by means of waiting lists or appointment systems, so capacity decisions are the most important tool to influence patient waiting times. Additional complexities result from the relatively small system size that characterizes an ED (i.e. a...

Many service systems are characterized by time-varying demand for service. For instance in an emergency department, patient arrival rates are usually not constant throughout the day. This arrival process is stochastic, but nonetheless predictable to some extent (a daily pattern can often be distinguished). However, this feature can severely complic...

In many inventory management systems, some kind of substitution flexibility exists, meaning that a substitute or more flexible item can be used (at an additional cost) when the preferred product stocks out. Through the use of substitution flexibility, we can take advantage of the risk pooling effect on the flexible item. Since risk pooling reduces...

This paper studies the optimal design of an inventory system with “one-way substitution” , in which a high-quality (and hence, more expensive) item fulfills its own demand and simultaneously acts as backup safety stock for the (cheaper) low-quality item. Through the use of a discrete time Markov model we analyze the effect of one-way substitution i...

This paper presents an integrated production inventory model that enables to capture the tradeoffs between average inventory, production capacity and customer service level in a semiprocess industry setting. The model includes different features that are specific for such a setting, such as differences in reactor yield and quality requirements acro...

Given the increased pressure on short delivery lead times, minimizing customer order throughput times is an important objective in warehousing operations. Warehouse managers can influence the responsiveness of their system through a number of controls, such as the order batching policy, the capacity of the picking and sorting operations, and the pi...

Lean management is widespread but theoretical models that scientifically substantiate the lean practice are scarce. We show how queuing models of manufacturing systems and supply chains underpin the practice of Lean. Two quantitative performance models which relate the system parameters with the system performance in terms of lead time and throughp...

The paper presents a robust operational planning procedure for a make-to-order environment. The procedure successively takes into account order deadlines (lead time offsetting process), allows for capacity flexibility and aims at minimizing set-ups. The novelty of the procedure lies in the use of realistic lead time estimates (including capacity lo...

This article examines the relationship between risk, return, skewness, and utility-based preferences. Examples are constructed showing that, for any commonly used utility function, it is possible to have two continuous unimodal random variables X and Y with positive and equal means, X having a larger variance and lower positive skewness than Y, and...

The planning and decision support capabilities of the manufacturing planning and control system, which provides the core of any enterprise resource planning package, can be enhanced substantively by the inclusion of a decision support module as an add-on at the midterm planning level. This module, called advanced resource planning (ARP), provides a...

This article proposes a supporting framework for the implementation of the material control system POLCA (paired-cell overlapping loops of cards with authorization). The POLCA system is particularly appropriate for environments that involve highly variable demand and large product variety, which force small batch (or even one-of-a-kind) production....

Our research focuses on improving the output rate of a single product CONWIP system. The system has zero intermediate buffer capacity but may use time buffers which result from minimum and maximum processing times on each workstation. The transport between the different workstations is accomplished by a single resource (a bridge crane), constitutin...

In this paper, we show that the planning and decision-support capabilities of the MPC (Manufacturing Planning and Control) system, which forms the core of any ERP (Enterprise Resource Planning) package, may be substantively enhanced by including a Decision Support Module (DSM) as an add-on at the midterm planning level. This DSM, called Advanced Re...

This paper describes two experiments exploring the potential of the Kriging methodology for constrained simulation optimization. Both experiments study an (s, S) inventory system with the objective of finding the optimal values of s and S. The goal function and constraints in these two experiments differ, as does the approach to determine the optim...

This paper studies the impact of management policies, such as product allocation and campaign sizing, on the required size of the finished goods inventories in a multi-product multi-reactor batch process. Demand, setup and batch processing times for these products are assumed to be stochastic, and the inventory buffer for every product type needs t...

In many warehouses, customer orders are batched to profit from a reduction in the order picking effort. This reduction has to be offset against an increase in sorting effort. This paper studies the impact of the order batching policy on average customer order throughput time, in warehouses where the picking and sorting functions are executed separa...

The objective of this article is to derive the density function and cumulative distribution function for random variables which may be written as the sum of independent (either identical or non-identical) zerotruncated Poisson random variables. The obtained expressions may be particularly useful for modelling purposes, especially in view of linking...

Current supply chain policies are increasingly influenced by the lean philosophy, which calls for increased synchronization and smaller but more frequent shipments between the supply chain partners. One of the strategies used in this respect is lot splitting, i.e. the splitting of a single order quantity among multiple deliveries. Lot splitting is...

This paper presents analytical expressions for estimating average process batch flow times through a stochastic manufacturing system with overlapping operations. It is shown that the traditional queueing methodology cannot be directly applied to this setting, as the use of the overlapping operations principle causes the arrival process of sublots a...

The impact of transfer batching (also referred to as lot splitting) on the performance of flowshops has received widespread attention in the literature. Most papers have emphasized the usefulness of lot splitting in cutting down average flow times, as it enables the overlapping of operations at different stages of the flowshop. However, as most ana...

This paper focuses on modelling the impact of lot splitting on average process batch flow times, in a two-stage stochastic flowshop. It is shown that the traditional queueing methodology for estimating flow times cannot be directly applied to a system with lot splitting, as the arrival process of sublots at the second stage is not a renewal process...

We propose a supporting framework for the material control system Paired-cell Overlapping Loops of Cards with Authorization (POLCA), which combines the advantages of push systems and pull systems. In our load based version of the POLCA control system, we rely on a multiproduct, multi-machine queueing network to determine release authorizations and...

In this paper, we study the decision problem of a retailer, who wants to optimize the amount of shelf inventory of a particular product, given that the demand for the product is stochastic and replenishment lead times (from the store’s stockroom to the shelf) are negligible. The shelf inventory is managed according to a (0,B*)-inventory policy: whe...

This paper describes a model for minimizing total costs in a single-product, deterministic flow shop with overlapping operations in terms of the sublot size used. Three types of costs are considered: the inventory holding costs, the transportation costs and the so-called “gap costs” which may result from the intermittent idling of machines between...

In this paper, we consider the problem of estimating the average and variance of replenishment lead time in a multi-product multi-reactor batch process with a base-stock inventory policy. The methodology used for this purpose is queueing theory. Although the body of knowledge on queueing theory has primarily been adopted in discrete manufacturing p...

In this paper we analyze how a nuclear magnetic resonance scanner can be managed more efficiently, simultaneously improving patient comfort (in terms of total time spent in the system) and increasing availability in case of emergency calls. By means of a superposition approach, all relevant data on the arrival and service process of different patie...

It is generally known that the mechanism of overlapping operations allows to reduce the average job lead times in a production system. However, a machine may remain idle between two consecutive sublots belonging to the same job, due to a difference between its own setup and processing times and those of the preceding machine. These idle times are r...

In practice, sequence changes during production often occur due to material shortages, rush jobs, etc. Rescheduling, expediting and de-expediting are common practice on the shop floor, but in many cases it is less clear to which extent such control policies influence throughput times. The aim of this paper is to study the impact of sequence changes...

It is now widely accepted that both large and small lot sizes can cause long lead times and consequently bad customer service in terms of late deliveries. The impact of the lot size on the lead time consists of a convex relationship, implying an optimal lot size minimising average lead time. In order to educate people with this undoubtedly correct,...

The performance of a real-life production system (as measured by, for instance, the average flow time of parts or the average work-in-process level) is impacted by a range of managerial decisions, among which the product mix being produced in the shop, the layout of the shop, and the lot sizing policies (also called batching policies) used on the s...