Date on Master's Thesis/Doctoral Dissertation


Document Type

Doctoral Dissertation

Degree Name

Ph. D.


Industrial Engineering

Degree Program

Industrial Engineering, PhD

Committee Chair

Gue, Kevin

Committee Co-Chair (if applicable)

Bai, Lihui

Committee Member

Bai, Lihui

Committee Member

Gentili, Monica

Committee Member

Pop, Dan

Author's Keywords

gridhub; grid-based; high-density; material handling system


In the past twenty years, the share of e-commerce has increased (FRED, 2019). Since more and more activities, such as picking and sorting customers’ orders, are done in warehouses, high efficiency warehouses are in demand. Furthermore, the efficiency of warehouses is related to customer satisfaction (Colla and Lapoule, 2012). Storage systems are key components in warehouses, which are related to the efficiency of warehouse operations. In this dissertation, we address an automatic puzzle-based storage system under decentralized control. We call this system GridHub. GridHub meets standards of Industry 4.0 (Lasi et al., 2014), and it features high throughput with parallel order processing. In the first part of this research, we describe a GridHub which can handle unit-sized items; that is, one box only occupies one conveyor module. The GridHub is capable of moving boxes in all cardinal directions. It can complete multiple material handling tasks, such as sorting, sequencing, retrieving, and storing without changing the control algorithms. To move the active boxes to their targets, we developed a decentralized control algorithm to arrange box movements. The algorithms are executed by conveyor modules cyclically, and all actions in the execution process are one iteration of the algorithm. There are three phases in one iteration (assess, negotiation and convey), and several steps consist of one phase. The conveyor modules execute the algorithm simultaneously and synchronize at every step. The goal of the control algorithms is to move active boxes into their immediate destinations, and the key idea of the algorithm is to move away other boxes for the active boxes through message passing process. Negotiation behaviors are patterns of action generated by the conveyor modules while executing the algorithms. We describe these behaviors and explain how they affect the transfer process of active boxes. Some of those behaviors and other actions, which can prevent the transferring processes of boxes, are listed and discussed. These actions are related to deadlock and livelock in the GridHub. We prove that GridHub is deadlock free, and it is also livelock free under certain conditions. In the second part of this research, we extend the unit-sized GridHub by enabling it to handle non-unit-sized boxes meaning every box can occupy more than one conveyor module. We name the new GridHub the NU GridHub. The control algorithms of the NU GridHub are developed based on the unit-sized GridHub’s algorithms by adding new rules. Performance of the NU GridHub is also measured and discussed. GridHub is the first grid-based material handling system to offer four-way movement of stored items with a rich set of material handling task – storage, retrieval, sorting, and sequencing. GridHub is also the first grid-based system to implement a decentralized control algorithm based on “nested attempts,” a feature the guarantee deadlock free operation. Finally, the NU GridHub is the first grid-based solution to handle bigger boxes, which have not been done for a grid-based system under the virtual aisle method.