About

Sigrun Andradottir is a Professor in the H. Milton Stewart School of Industrial and Systems Engineering at Georgia Tech. Her research interests focus on simulation, applied probability, and stochastic optimization. She has developed expertise in these areas, contributing to the fields of applied probability and simulation within industrial and systems engineering. Sigrun Andradottir earned her B.S. in Mathematics from the University of Iceland in 1986, followed by an M.S. in Statistics from Stanford University in 1989, and a Ph.D. in Operations Research from Stanford University in 1990. After completing her doctorate, she joined the faculty of the University of Wisconsin – Madison in 1990 and later moved to Georgia Tech in 1995. Throughout her career, she has been affiliated with professional organizations such as INFORMS and Operations Research, reflecting her active engagement with the research community in her areas of specialization.

Research topics

• Computer Science
• Artificial Intelligence
• Mathematics
• Mathematical optimization
• Machine Learning
• Operations research
• Data Mining
• Nursing
• Economics
• Medicine
• Microeconomics
• Family medicine

Selected publications

• Pooling in tandem queueing networks with non-collaborative servers

  UNC Libraries · 2026-04-03

  articleOpen access1st authorCorresponding
  PublisherDOI

• Optimal resident physician supervision when patient care is non-preemptive

  Annals of Operations Research · 2026-05-21

  articleOpen access1st author

  Abstract This paper focuses on the assignment of attending physicians between the residents they supervise and their own responsibilities. Unlike prior work that assumes patient care can be interrupted at any time, we consider the more realistic and technically challenging situation when patient care is non-preemptive. Under the assumption that a holding cost is incurred when residents and patients wait for a conference with the attending physician and that rewards are earned whenever the attending physician completes a task (on his own or with his residents), we completely characterize the allocation of the attending physician that maximizes the long-run average profit. Furthermore, we show that the optimality conditions are simple thresholds on the holding cost. We also discuss how the optimal allocation of the attending physician differs from those in systems with preemptions. In particular, we show that the main difference occurs when the holding cost is high and there is no resident waiting for a conference. In this case, the attending physician waits for a resident to be ready for consultation in the non-preemptive case, whereas he works on his own responsibilities in the preemptive model. We conclude with a study of various extensions of our attending physician and residents interaction model and show that the structure of the optimal policy remains the same in these more general settings.

  PublisherOA PDFDOI

• Collaboration versus Specialization in Service Systems with Impatient Customers

  arXiv (Cornell University) · 2026-01-22

  preprintOpen access

  We study tandem queueing systems in which servers work more efficiently in teams than on their own and customers are impatient in that they may leave the system while waiting for service. Our goal is to determine the server assignment policy that maximizes the long-run average throughput. We show that when each server is equally skilled at all tasks, the optimal policy has all the servers working together at all times. We also provide a complete characterization of the optimal policy for Markovian systems with two stations and two servers when each server's efficiency may be task dependent. We show that the throughput is maximized under the policy which assigns one server to each station (based on their relative skill at that station) unless station 2 has no work (in which case both servers work at station 1) or the number of customers in the buffer reaches a threshold whose value we characterize (in which case both servers work at station 2). We study how the optimal policy varies with the level of server synergy (including no synergy) and also compare the optimal policy for systems with different customer abandonment rates (including no abandonments). Finally, we investigate the case where the synergy among collaborating servers can be task-dependent and provide numerical results.

  PublisherDOI

• Collaboration versus Specialization in Service Systems with Impatient Customers

  ArXiv.org · 2026-01-22

  articleOpen access

  We study tandem queueing systems in which servers work more efficiently in teams than on their own and customers are impatient in that they may leave the system while waiting for service. Our goal is to determine the server assignment policy that maximizes the long-run average throughput. We show that when each server is equally skilled at all tasks, the optimal policy has all the servers working together at all times. We also provide a complete characterization of the optimal policy for Markovian systems with two stations and two servers when each server's efficiency may be task dependent. We show that the throughput is maximized under the policy which assigns one server to each station (based on their relative skill at that station) unless station 2 has no work (in which case both servers work at station 1) or the number of customers in the buffer reaches a threshold whose value we characterize (in which case both servers work at station 2). We study how the optimal policy varies with the level of server synergy (including no synergy) and also compare the optimal policy for systems with different customer abandonment rates (including no abandonments). Finally, we investigate the case where the synergy among collaborating servers can be task-dependent and provide numerical results.

  PublisherOA PDF

• Indifference-Zone Relaxation Procedures for Finding Feasible Systems

  ArXiv.org · 2025-09-02

  preprintOpen access

  We consider the problem of finding feasible systems with respect to stochastic constraints when system performance is evaluated through simulation. Our objective is to solve this problem with high computational efficiency and statistical validity. Existing indifference-zone (IZ) procedures introduce a fixed tolerance level, which denotes how much deviation the decision-maker is willing to accept from the threshold in the constraint. These procedures are developed under the assumption that all systems' performance measures are exactly the tolerance level away from the threshold, leading to unnecessary simulations. In contrast, IZ-free procedures, which eliminate the tolerance level, perform well when systems' performance measures are far from the threshold. However, they may significantly underperform compared to IZ procedures when systems' performance measures are close to the threshold. To address these challenges, we propose the Indifference-Zone Relaxation (IZR) procedure, IZR introduces a set of relaxed tolerance levels and utilizes two subroutines for each level: one to identify systems that are clearly feasible and the other to exclude those that are clearly infeasible. We also develop the IZR procedure with estimation (IZE), which introduces two relaxed tolerance levels for each system and constraint: one matching the original tolerance level and the other based on an estimate of the system's performance measure. By employing different tolerance levels, these procedures facilitate early feasibility determination with statistical validity. We prove that IZR and IZE determine system feasibility with the desired probability and show through experiments that they significantly reduce the number of observations required compared to an existing procedure.

  PublisherOA PDFDOI

• Optimal Control of Queueing Systems with Error-Prone Servers

  Stochastic Systems · 2025-08-11 · 1 citations

  articleOpen access

  Consider a Markovian tandem line with finite intermediate buffers and an equal number of stations and servers. Servers are flexible but noncollaborative, so that a job can be processed by at most one server at any time. When a job is being processed, it can be damaged and wasted depending on the proficiency of the server. We identify the dynamic server assignment policy that maximizes the long-run average throughput of the system with two stations and two servers. We find that the optimal policy is either a single or a double threshold policy on the number of jobs in the buffer, where the thresholds depend on the service rates and defect probabilities of the two servers at the two stations. For larger systems, we show that the optimal policy may involve server idling and that improving the service rate at any station is always beneficial. Finally, we propose heuristic server assignment policies motivated by experimentation for small systems with finite buffers and analysis of larger systems with infinite buffers. Numerical results suggest that our heuristics yield near-optimal performance. Funding: This research was supported by the National Science Foundation [Grants CMMI-1536990 and CMMI-2127778]. S. Andradóttir was also supported by the National Science Foundation [Grant CMMI-2348409].

  PublisherDOI

• Optimal pricing and information sharing strategies in a single-server queue

  Queueing Systems · 2025-12-08

  article
  PublisherDOI

• Optimal server control with Two Customer Classes and Classification Errors

  European Journal of Operational Research · 2025-12-11

  articleOpen access1st author

  • Customer misclassification is a common phenomenon in many applications. • We provide analytical models of service systems with customer misclassification. • We identify the optimal allocation of specialists in systems with customer misclassification. • We investigate how the long-run average profit depends on the misclassification probability. • We identify under what conditions it is more profitable to serve customers with or without service continuity. We consider a Markovian queueing system with two types of customers (basic and advanced) and two types of servers (regular and specialist) in the presence of customer classification errors. We assume that there are always both types of customers waiting for service. When an advanced customer is misclassified as a basic customer, he needs to be served by a specialist after being served by a regular server. Our objective is to determine the dynamic assignment of the specialists between advanced and misclassified customers that maximizes the long-run average profit. We consider two versions of the problem that differ depending on whether the misclassified customers experience service continuity (the regular servers stay with misclassified customers while they wait for specialists, preventing the regular servers from serving other basic customers) or not (the regular servers continue serving other basic customers while misclassified customers wait for specialists). For both versions of the problem, we first characterize the optimal assignment of the specialists and then investigate how the optimal long-run average profit depends on the misclassification probability. We provide examples of systems where the optimal long-run average profit is not monotone in the misclassification probability, which is counter intuitive as one would expect misclassification to have a negative impact on system performance. We conclude our analysis by identifying under what conditions it is more profitable to serve customers with or without service continuity.

  PublisherDOI

• Pricing in Queues with Abandonments: Optimal Policies and Practical Heuristics

  ArXiv.org · 2025-05-15

  preprintOpen access

  We investigate the optimal pricing strategy in a service-providing framework, where customers can leave the system prior to service completion. In this setting, a price is quoted to an incoming customer based on the current number of customers in the system. When the quoted price is lower than the price the incoming customer is willing to pay (which follows a fixed probability distribution), then the customer joins the system and a reward equal to the quoted price is earned. A cost is incurred upon abandonment and a holding cost is incurred for customers waiting to be served. Our goal is to determine the pricing policy that maximizes the long-run average profit. Unlike traditional queueing systems without abandonments, we show that the optimal quoted prices do not always increase with the queue length in this setting. We fully characterize the possible structure of the optimal dynamic pricing policy and provide conditions guaranteeing that the optimal policy is increasing in the number of customers in the system. Moreover, we introduce two heuristics that simplify the optimal dynamic pricing policy. Both heuristics admit customers until the number of customers in the system reaches a certain threshold. The cutoff-static policy charges all admitted customers a fixed price while the two-price policy charges one price when the arriving customer can enter service immediately and another price if the customer needs to wait. By selecting the price(s) and threshold that maximize the long-run average profit, both heuristics achieve near optimality in general and the two-price policy provides more robustness compared to the cutoff-static policy.

  PublisherOA PDFDOI

• Finding Feasible Systems in the Presence of a Probability Constraint

  2024-12-15

  article

  We consider the problem of determining feasible systems among a finite set of simulated alternatives with respect to a probability constraint, where observations from stochastic simulations are Bernoulli distributed. Most statistically valid procedures for feasibility determination consider constraints on the means of normally distributed observations. When observations are Bernoulli distributed, one can still use the existing procedures by treating batch means of Bernoulli observations as basic observations. However, achieving approximate normality may require a large batch size, which can lead to unnecessary waste of observations in reaching a decision. This paper proposes a procedure that utilizes Bernoulli-distributed observations to perform feasibility checks. We demonstrate that when the observations are Bernoulli distributed, our procedure outperforms an existing feasibility determination procedure that was developed for a constraint on normally distributed observations.

  PublisherDOI

Recent grants

• Queueing Systems with Server Synchronization

  NSF · $350k · 2015–2020

• Resource Flexibility for Uncertainty Mitigation in Production and Service Systems

  NSF · $360k · 2009–2014

• Improving the Performance of Queueing Systems through Cross-Training

  NSF · $318k · 2000–2005

Frequent coauthors

• Hayriye Ayhan

  Georgia Institute of Technology

  27 shared

• David Goldsman

  Georgia Institute of Technology

  13 shared

• Douglas G. Down

  McMaster University

  13 shared

• Seong‐Hee Kim

  Sungkyunkwan University

  12 shared

• Nilay Tanık Argon

  University of North Carolina at Chapel Hill

  9 shared

• Andrei A. Prudius

  Bloomberg (United States)

  7 shared

• Teunis J. Ott

  Los Alamos National Laboratory

  6 shared

• Mehdi Hoseyni-Nasab

  Rutgers, The State University of New Jersey

  6 shared