Functional Analysis Of Food Distribution

Author : Osama Alotaik
Publisher :
ISBN 13 :
Total Pages : pages
Book Rating : 4.:/5 (127 download)

DOWNLOAD NOW!

Download or read book Equitable Food Distribution-Modeling and Analysis of Food Banks Supply Chains Using Machine Learning and Recurrent Optimization-based Simulation written by Osama Alotaik and published by . This book was released on 2021 with total page pages. Available in PDF, EPUB and Kindle. Book excerpt: Food banks are non-profit organizations that collect and distribute food to hunger-relief charities. Food banks act as food storage and distribution depots for smaller front line agencies; they usually do not themselves give out food directly to people struggling with hunger. Food banks have a moral and social imperative to ensure that food is distributed equitably to those in need within their service area. Equitable food distribution is also a requirement from the Feeding America organization. So, the important issues that need to be addressed are: how can food banks allocate food in an equitable manner? What factors inhibit them from achieving the goal of equity? How can these factors be overcome? We consider the context of delivery orders being placed by food agencies, that is the food bank collects the orders and then makes food allocation decisions under capacity constraints. We divided this research into three sub-problems. The first sub-problem addresses a gap in the literature with respect to factors affecting equity in the context of food assistance. It is often assumed that equity could be measured by either spatial factors or the size of poor population. When using spatial factors and food accessibility measures, under-served areas are identified to be those with low accessibility to food agencies. On the other hand, using the size of poor population ignores potential disparities in terms of food accessibility. So, the questions we ask here are: how do demographics, spatial accessibility, demand patterns, and food distribution patterns affect the equity of food distribution? What are the possible biases that could emerge from the interaction of these factors and cause food to be distributed inequitably? Upon examining the characteristics of under-served areas in the first part of this research, it was found that the receiving capacity of food agencies could be an inhibiting factor from achieving equitable food distribution. Receiving capacity is the ability of food agencies to receive and store food. So, the question of the second sub-problem is: does the receiving capacity of food agencies have an impact on equity? Viewing the problem as a resource allocation problem where the resources are the receiving capacity of food agencies, the specific question we ask is: will increasing the receiving capacity of under-served areas improve the equity in food distribution? The third part of this research briefly addresses the problem of network design for food bank supply chains. Although the proposed approach in the second part does optimize the network design of food banks, it is specifically geared towards improving equity of food distribution. In this part, the problem of network design is addressed in the general sense. Viewing the problem as a facility location problem, we examine how food banks can design their network of food agencies under various objectives such as minimizing the total distance between agencies and populations centers; maximizing population coverage; and minimizing the number of agencies while maximizing population coverage. To address the first sub-problem, a machine learning-based framework is developed and utilized to understand the factors that could cause inequitable distribution of food and to evaluate the status of food assistance in a population. Unsupervised learning, namely clustering, is used to carry out the framework. We illustrate how to use publicly available data in doing so. We showed that identifying under-served or over-served areas based solely on spatial factors, or the size of poverty populations could be misleading. We've shown that the disparity between rural and urban areas should be examined using additional factors such as food distribution patterns. We explained that limited storage resources could possibly be the cause for such disparity. To address the second sub-problem, a novel heuristic modeling approach was developed to utilize the output of the first phase for the purpose of optimizing resource allocation in a supply chain. The approach is composed of two main components: a mathematical model that uses the clustering output labels to make resource allocation decisions. The second component is a Recurrent Optimization-based Simulation (ROS) model that is mainly used for experimentation and to validate the resource allocation decisions. It was found that increasing the receiving capacity of under-served block groups provided significant improvements in terms of equity of food distribution. Lastly, a suite of integer programming models were briefly introduced to optimize the design of food banks networks. Design of food banks supply chains is found to be essential for achieving the goal of equity. Therefore, we briefly presented some of the common models used for facility location problems and showed how they could be applied in the context of food banks. We found that existing models for network design are mostly suitable for efficiency related objectives such as minimizing total costs; but are not necessarily suitable for the objective of equity. In this thesis we collected data from several sources and identified and extracted the data we required. We did painstaking exploratory analysis to make sure the data is verifiable and validated.