Seminar: Orbit Size Dependent Return Flow Estimation for Closed Loop Supply Chains

SEMINAR 

DEPARTMENT OF INDUSTRIAL ENGINEERING 

Orbit Size Dependent Return Flow Estimation for Closed Loop Supply Chains

Seval Ata Demirkan,

Invent Analytics 

Abstract:

In this study, a method is developed to estimate return flow based on the number of products in use (orbit) for the mature phase of a product’s life cycle within a partially observable hybrid production system featuring an unobservable product disposal process. The mature phase is characterized by stationary demand and return processes, with the latter depending on the orbit size, while the production control mechanism also needs to be considered. Accurately estimating the orbit size is valuable for estimating product returns, which can impact production decisions. Therefore, it is aimed to estimate the orbit size dynamically considering all possible event occurrences such as demand arrivals, product returns and product disposals. For this purpose, a simulation model of the hybrid production system that utilizes partial information obtained from product returns is built. Bayesian Estimation (BE) and Moving Average (MA) methods are adapted to the model, and a novel approach called the Two-State Dependent Filtering Algorithm (TSD-FA) is introduced. This approach extends the traditional Hidden Markov Model by allowing observations to depend on the last two hidden states. An approximation to TSD-FA, denoted as ATSD-FA, is also proposed, which limits the number of observable events in a review period. These estimation methods are evaluated under both orbit-dependent and independent inventory control policies. Furthermore, the value of partial information is highlighted by evaluating an alternative estimation method that ignores observations. Numerical results demonstrate that TSD-FA and ATSD-FA outperform BE and MA in terms of accuracy and robustness.

Keywords: Hybrid Production Systems, Return Flow Estimation, State-Dependent Filtering, Hidden Markov Models, Closed Loop Supply Chains

Short Bio: 

Seval Ata Demirkan is a graduate of Boğaziçi University, where she earned her Bachelor's and Master's degrees in Industrial Engineering in 2015 and 2018, respectively. She completed her PhD in Industrial Engineering at Boğaziçi University in October 2024, with research focused on supply chains, particularly reverse supply chain mechanisms. Currently, Seval works as a Data Scientist at Invent Analytics, applying her academic expertise in supply chains to real-world challenges.

All interested are cordially invited.  

DATE:  March 6, 2026 

TIME:  Friday, 15:00

ROOM: Engineering Building, M3100