Stochastic modeling and analysis of automated Intra-logistics systems

Abstract

Logistics companies are adopting automated technologies for material handling operations in intralogistics facilities, such as sea container terminals and distribution warehouses, to improve operational efficiency and meet customer expectations of higher service levels. While automated technologies are promising for higher operational efficiency, the full benefit of automation is leveraged by optimizing system design parameters. In this thesis, we develop stochastic models for intra-logistics systems, which help in analyzing system performance under various parameter settings and allow optimization of design parameters. The thesis comprises three essays. The first essay deals with stochastic modeling of parallel process flows in intra-logistics systems, where different resources involve simultaneous (parallel) operations while processing a job. We propose an efficient modeling approach for parallel process flows using two-phase servers. We develop a closed queuing network model to estimate system performance measures. For solving the resulting closed queuing network consists of two-phase servers, we develop two solution methods: an extended approximate mean value analysis and a network aggregation dis-aggregation approach. We illustrate the proposed modeling approach for two intra-logistic systems: a sea container terminal with automated guided vehicles and a shuttle-based compact storage system. In the second essay, we propose a novel solution method for multi-stage semi-open queuing networks (SOQNs) that are widely used to measure the performance of manufacturing and warehousing systems. While there are several methods available for solving single-stage SOQNs, solution methods for multi-stage SOQNs are limited. We propose a method that decomposes the multi-stage SOQN into multiple single-stage SOQNs, estimates the job departure process information from upstream singlestage SOQNs and links them together to obtain the performance of the original multi-stage SOQN. We demonstrate the efficacy of the proposed approach using a case study on a multi-tier shuttle-based compact storage system and benchmark our results with an existing approach. In the third essay, we study the performance trade-offs between the technology choices (lift-automated guided vehicles vs. automated guided vehicles) for robotic transport vehicles in automated container terminals under various design parameter settings. We develop a stylized SOQN, which contains two-phase servers and finite capacity queues, and solve it using a novel network decomposition method.