Research

Contributing to building a sustainable society drives Masoud’s research. Efficient and effective design of business processes is one of the pillars of economic prosperity in a sustainable society. He combines scientific ambition and rigor with practical relevance in his research.

A review of his publications in top-tier peer-reviewed research outlets as well as those in the pipeline is listed below.

Masoud Mirzaei, Ph.D. dissertation. ‘Advanced Storage and Retrieval Policies in Automated Warehouses’, Rotterdam School of Management (2020).

Companies operating close to the rural or dense industrial areas face a high capital cost. One line of his current research focuses on the analytical model and operational design of high-density storage systems.

High-density storage systems have a high space utilization due to a lack of transportation aisles. Here, throughput capacity and response time are the critical performance measures of the system. In this research where items are stored on mobile shuttles, he proposes efficient unit-load handling methods that improve the throughput capacity of the storage and retrieval system and reduce the order picking makespan. He formulates a mathematical model and a heuristic solution to find the optimal simultaneous retrieval time for two loads and multiple loads, respectively. He uses the simulation for validation. The model and results are discussed in detail in his published paper as:

Mirzaei, Masoud; De Koster, René; Zaerpour, Nima: Modelling Load Retrievals in Puzzle-based Storage Systems. International Journal of Production Research 55: 6423-6435 (2017).

DOI: 10.1080/00207543.2017.1304660.

The complexity of the optimization problem increases when the high-density storage systems have a few open locations available while operating. This situation can lead to an increased throughput capacity but requires specific design and control. Masoud is currently developing exact retrieval methods for high-density systems where multiple requested loads move simultaneously using a few open locations in the system. Applications can be found in urban distribution centers and automated parking lots.

Mirzaei, Masoud: Design, Modeling, and Analysis of High-Density Storage Systems.

To be submitted for publication in spring 2021.

Digitalization has made a considerable amount of data available for improving the design and planning of the processes and optimize operational performance. Another line of Masoud’s research uses historical data to redesign intralogistics operations in warehouses and fulfillment centers, namely the storage strategies and order picking policies.

His research on storage and retrieval policies derives analytical insights from historical customer demand data to design efficient storage and retrieval strategies in automated warehouses (e.g., AS/RS and Amazon robotics). The aim is to achieve a shorter order picking travel time at a lower cost. A mixed-integer program models the concurrent clustering of correlated products and allocation of clusters to storage space, considering both turnover frequency and product affinity in historical demand data. A scenario-based analysis of real data evaluates the effect of demand parameters on the performance of the model. He developed a heuristic that solves large instances. The model is validated using a second dataset of thousands of products and orders. The results suggest that a cluster-based storage strategy may reduce the order picking time when the affinity between products is high, and the turnover frequency curve is not highly skewed. The detail of the model and the numerical analysis are discussed in the following paper:

Mirzaei, Masoud; Zaerpour, Nima; De Koster, René: The Impact of Integrated Cluster-based Storage Assignment on Parts-to-Picker Warehouse Performance. Transportation Research- Part E, 146: February 2021, 102207.

DOI: https://doi.org/10.1016/j.tre.2020.102207.

Splitting the inventory of each product and spreading them over the warehouse make them accessible for order pickers, thus helps reducing order picking time. The next paper develops a mixed-integer program for designing a correlated storage assignment that disperses the inventory optimally through the warehouse. Masoud develops an efficient construction and improvement heuristic to solve large scale problems. He also uses statistical analysis to evaluate the relation of order picking performance and order profile. This experiment is based on a big dataset from a Dutch retailer in the numerical analysis. Results show a correlated dispersed storage assignment leads to shorter retrieval time compared to common storage policies for sufficiently large order sizes. Correlation of products in customer demand has a key role in the performance of the model. This paper is under review at the IISE Transactions:

Mirzaei, Masoud; Zaerpour, Nima; De Koster, René: How to Benefit from Order Data: Correlated Dispersed Storage Assignment in Robotic Warehouses. (2020, European Journal of Operational Research).

Under review.

Batch order picking is a common method to reduce travel distance and improve the performance of fulfillment centers. Orders are batched based on picker routing, the capacity of the pick station, and most importantly the storage strategy used in the warehouse. Common storage strategies, such as random, turnover-based and similarity-based assignments, optimize based on the information at the order level. They do not consider batch picking, therefore their design results in underperformance of the batch order picking process. This research takes a backward design approach from batch picking to an optimal storage strategy. This paper 1) evaluates the effect of batching choices on the design correlation, 2) models optimal storage strategies for batch picking to reduce the order picking time and increase the throughput capacity of the system, 3) optimizes the batch picking based on the storage strategy design, and 4) compares the analytical results on the performance of these strategies compared to traditional ones.

Mirzaei, Masoud; Adan, Ivo; De Koster, René: A Backward Design: Storage Strategies for Batch Order Picking.

To be submitted for publication in summer/fall 2021.

Hub and spoke networks are used to switch and transfer commodities between terminal nodes in distribution systems at minimum cost and/or time. Previous studies consider only quantitative parameters such as cost and time to find the optimum location. Often the critical role of qualitative parameters like quality of service, zone traffic, environmental issues, capability for development in the future and etc., that are critical for decision-makers have not been incorporated into models. In many real-world situations, qualitative parameters are as much important as quantitative ones. This paper presents a hybrid approach to the p-hub center problem in which the location of hub facilities is determined considering both parameters simultaneously.

Bashiri, Mahdi; Mirzaei, Masoud, Randall, Marcus: Modeling Fuzzy Capacitated p-hub Center Problem and a Genetic Algorithm Solution. Applied Mathematical Modelling 37: 3513–3525 (2013).

DOI: 10.1016/j.apm.2012.07.018.

Other publications include a booklet and conference proceedings listed below.

Mirzaei, Masoud; Bashiri, Mahdi: Ant Colony Optimization: Concepts, Algorithms and Programming. Tehran, the Commerce Printing and Publishing Company (2010).

– in Persian

Mirzaei, Masoud; Bashiri, Mahdi: Multiple Objective Multiple Allocation Hub Location Problem. The 40th International Conference on Computers & Industrial Engineering, Awaji, Japan, IEEE (2010).

DOI: 10.1109/ICCIE.2010.5668249.

Mirzaei, Masoud; Asadi, Hashem: Modeling Capacitated Hub Median Problem. 3rd International Conference in Operations Research, Iran (2010).

COI: ICIORS03_298, – in Persian.

Bashiri, Mahdi; Mirzaei, Masoud: Hybrid Fuzzy Capacitated Hub Center Allocation Problem with both Qualitative and Quantitative Variables. World Applied Sciences Journal, 5: 507–516 (2008).

Read online.