Research Outputs

Permanent URI for this communityhttps://hdl.handle.net/20.500.14288/2

Browse

Filters

2017 - 201912020 - 20235

Advanced Search

Filter by

Settings

Sponsors: search.filters.sponsors.EUSubject: search.filters.subject.Operations research

Search Results

Now showing 1 - 6 of 6
  • Thumbnail Image
    PublicationOpen Access
    A machine learning approach for implementing data-driven production control policies
    (Taylor _ Francis, 2021) Department of Business Administration; N/A; Tan, Barış; Khayyati, Siamak; Faculty Member; Department of Business Administration; College of Administrative Sciences and Economics; Graduate School of Sciences and Engineering; 28600; N/A
    Given the extensive data being collected in manufacturing systems, there is a need for developing a systematic method to implement data-driven production control policies. For an effective implementation, first, the relevant information sources must be selected. Then, a control policy that uses the real-time signals collected from these sources must be implemented. We analyse the production control policy implementation problem in three levels: choosing the information sources, forming clusters of information signals to be used by the policy and determining the optimal policy parameters. Due to the search-space size, a machine-learning-based framework is proposed. Using machine learning speeds up optimisation and allows utilising the collected data with simulation. Through two experiments, we show the effectiveness of this approach. In the first experiment, the problem of selecting the right machines and buffers for controlling the release of materials in a production/inventory system is considered. In the second experiment, the best dispatching policy based on the selected information sources is identified. We show that selecting the right information sources and controlling a production system based on the real-time signals from the selected sources with the right policy improve the system performance significantly. Furthermore, the proposed machine learning framework facilitates this task effectively.
  • Placeholder
    PublicationMetadata only
    Bounded rationality in clearing service systems
    (Elsevier, 2020) Department of Industrial Engineering; Canbolat, Pelin Gülşah; Faculty Member; Department of Industrial Engineering; College of Engineering; 108242
    This paper considers a clearing service system where customers arrive according to a Poisson process, and decide to join the system or to balk in a boundedly rational manner. It assumes that all customers in the system are served at once when the server is available and times between consecutive services are independently and identically distributed random variables. Using logistic quantal-response functions to model bounded rationality, it first characterizes customer utility and system revenue for fixed price and degree of rationality, then solves the pricing problem of a revenue-maximizing system administrator. The analysis of the resulting expressions as functions of the degree of rationality yields several insights including: (i) for an individual customer, it is best to be perfectly rational if the price is fixed; however, when customers have the same degree of rationality and the administrator prices the service accordingly, a finite nonzero degree of rationality uniquely maximizes customer utility, (ii) system revenue grows arbitrarily large as customers tend to being irrational, (iii) social welfare is maximized when customers are perfectly rational, (iv) in all cases, at least 78% of social welfare goes to the administrator. The paper also considers a model where customers are heterogeneous with respect to their degree of rationality, explores the effect of changes in distributional parameters of the degree of rationality for fixed service price, provides a characterization for the revenue-maximizing price, and discusses the analytical difficulties arising from heterogeneity in the degree of bounded rationality. (C) 2019 Elsevier B.V. All rights reserved.
  • Thumbnail Image
    PublicationOpen Access
    Optimal sales and production rollover strategies under capacity constraints
    (Elsevier, 2021) Schwarz, Justus Arne; Department of Business Administration; Tan, Barış; Faculty Member; Department of Business Administration; College of Administrative Sciences and Economics; 28600
    Firms regularly replace their old product generation by a newer generation to sustain and increase their market share and profit. The product rollover problem of deciding on the number of old products to be pre-produced before the introduction of the new generation, and then deciding on the prices, sales volumes, and production volumes of the old and the new generation during the introduction under capacity constraint is considered. Production capacity limitations are common during the introduction period of a new product. We provide the first study that examines how a production capacity constraint affects the optimal decisions. The optimal decisions for a deterministic period-based model are provided in closed-form. A single sales/production rollover strategy implies that the sales/production of the old generation is discontinued before introducing the new generation. With a dual sales/production rollover strategy, the old and the new generation are sold/produced simultaneously. Depending on the capacity shortage, there are two types of mitigation actions: (i) increasing the prices, (ii) changing the sales and/or production rollover strategies with pre-production while adjusting the prices accordingly. If the capacity is unlimited, aligned sales and production rollover strategies are always optimal. We establish the conditions under which limited capacity leads to a combination of a single production rollover with a dual sales rollover strategy. We show that the selection of optimal rollover strategies is non-monotone in the available capacity. This implies that a change in the rollover strategy in response to limiting capacity has to be revoked for more severe capacity shortages.
  • Placeholder
    PublicationMetadata only
    Pricing in a transportation station with strategic customers
    (Wiley, 2017) N/A; Department of Industrial Engineering; Department of Industrial Engineering; Department of Industrial Engineering; Manou, Athanasia; Canbolat, Pelin Gülşah; Karaesmen, Fikri; Faculty Member; Faculty Member; Faculty Member; Department of Industrial Engineering; College of Engineering; College of Engineering; College of Engineering; N/A; 108242; 3579
    We consider a transportation station, where customers arrive according to a Poisson process, observe the delay information and the fee imposed by the administrator and decide whether to use the facility or not. a transportation facility visits the station according to a renewal process and serves all present customers at each visit. We assume that every customer maximizes her individual expected utility and the administrator is a profit maximizer. We model this situation as a two-stage game among the customers and the administrator, where customer strategies depend on the level of delay information provided by the administrator. We consider three cases distinguished by the level of delay information: observable (the exact waiting time is announced), unobservable (no information is provided) and partially observable (the number of waiting customers is announced). in each case, we explore how the customer reward for service, the unit waiting cost, and the intervisit time distribution parameters affect the customer behavior and the fee imposed by the administrator. We then compare the three cases and show that the customers almost always prefer to know their exact waiting times whereas the administrator prefers to provide either no information or the exact waiting time depending on system parameters.
  • Placeholder
    PublicationMetadata only
    Risk-sensitive control of branching processes
    (Taylor and Francis inc, 2021) Department of Industrial Engineering; Canbolat, Pelin Gülşah; Faculty Member; Department of Industrial Engineering; College of Engineering; 108242
    This article solves the risk-sensitive control problem for branching processes where the one-period progeny of an individual can take values from a finite set. the decision maker is assumed to maximize the expected risk-averse exponential utility (or to minimize the expected risk-averse exponential disutility) of the rewards earned in an infinite horizon. individuals are assumed to produce progeny independently, and with the same probability mass function if they take the same action. This article characterizes the expected disutility of stationary policies, identifies necessary and sufficient conditions for the existence of a stationary optimal policy that assigns the same action to all individuals in all periods, and discusses computational methods to obtain such a policy. are available for this article. See the publisher's online edition of IIE Transactions, datasets, Additional tables, detailed proofs, etc.
  • Placeholder
    PublicationMetadata only
    The digital twin synchronization problem: framework, formulations, and analysis
    (Taylor & Francis Inc, 2023) Matta, Andrea; Department of Business Administration; Tan, Barış; Department of Business Administration; College of Administrative Sciences and Economics
    As the adoption of digital twins increases steadily, it is necessary to determine how to operate them most effectively and efficiently. In this article, the digital twin synchronization problem is introduced and defined formally. Frequent synchronizations would increase cost and data traffic congestion, whereas infrequent synchronizations would increase the bias of the predictions and yield wrong decisions. This work defines the synchronization problem variants in different contexts. To discuss the problem and its solution, the problem of determining when to synchronize an unreliable production system with its digital twin to minimize the average synchronization and bias costs is formulated and analyzed analytically. The state-independent, state-dependent, and full-information solutions have been determined by using a stochastic model of the system. Solving the synchronization problem using simulation is discussed, and an approximate policy is proposed. Our results show that the performance of the state-dependent policy is close to the optimal solution that can be obtained with full information and significantly better than the performance of the state-independent policy. Furthermore, the approximate periodic state-dependent policy yields near-optimal results. To operate digital twins more effectively, the digital twin synchronization problem must be considered and solved to determine the optimal synchronization policy.