Approach to Fleet Contracting for Snow Plowing Operations Layek L. Abdel-Malek 1 ; Steven I. Chien, M.ASCE 2 ; Jay N. Meegoda, F.ASCE 3 ; and Haifeng Yu 4 Abstract: Ensuring a transportation system’ s operational efficiency and effectiveness is a challenge particularly for road maintenance agen- cies in winter times. Winter road maintenance involves snow plowing and chemicals/abrasives spreading, and both are performed by using a number of specific types of trucks. However, most departments of transportation do not maintain sufficient trucks and hence they may resort to contract additional equipment to make up for the shortfall. Based on the Newsvendor problem, this study develops a novel stochastic model to optimize the number of plows that should be contracted prior to a snow season to minimize the total cost, considering the frequency and distribution of various intensities of snow events, geometric condition of the roadways and traffic speeds. A solution methodology based on stochastic modeling is proposed and applied to a numerical example utilizing actual snow data. Additionally, a sensitivity analysis is performed to evaluate the impact of model parameters on the optimum solution. DOI: 10.1061/(ASCE)IS.1943-555X.0000202. © 2014 American Society of Civil Engineers. Author keywords: Winter road maintenance; Cost; Fleet size; Snow plowing; Newsvendor problem. Introduction Road maintenance in winter times involves many costly operations including spreading of chemicals and abrasives, snow plowing, loading snow into trucks, and hauling snow to disposal sites. In the United States, road maintenance operations consume over 2.3 billion dollars in direct cost each year [Federal Highway Administration (FHWA) 2011]. In recent years, new technologies in the applications of road weather information systems, weather forecasting services, and intelligent transportation systems have been implemented in many agencies to improve the process performance. An important goal of winter maintenance is to keep roads for safe passage. Snow falling on a paved surface may be removed by chemical, thermal, or mechanical techniques. Among the common practices to achieve reasonable traffic movement during inclement weather are applications of deicing materials and plowing. Under mounting pressure of high demand for improved winter safety and mobility under tight budgets, it is imperative for trans- portation agencies to seek the most cost-effective use of their re- sources. Hence, sound snow plowing models are desirable to determine the number of maintenance trucks needed to contract or call out for clearing designated road sections. Generally, agencies responsible for snow plowing contract a certain number of trucks prior to the winter season based on experiences or calcu- lation based on the worst case scenario of weather and traffic conditions. Nevertheless, the current practice may be neither economical nor necessarily rendering an improved performance. This is what motivates us to develop the approach presented in this paper. This paper considers the impact of under estimating and over estimating the needed number of plows for a forthcoming snow season. As will be seen in the literature review section, many articles have appeared in evaluation of road situation and the demand on snow plows. However, most of them have not suffi- ciently addressed various costs caused by over or under estimating the number of contracted plows, which could generate extra unneeded expenditures. The objective of this paper is to develop a model that minimizes the sum of aforementioned costs to assist state governments or transportation agencies in contracting the optimum number of snow plows needed and outsource when conditions necessitate. The approach in this study utilizes what is known in the liter- ature as the Newsvendor (or called Newsboy) problem. The di- lemma of the Newsvendor is that the night before one does not know exactly how the demand for the newspaper shall be on the following day. If the newsvendor orders many copies and there is low demand, he/she can salvage them as recycle but that will be lower than the purchasing cost. On the other hand if he/she orders smaller number of copies and the demand is large, he/she loses rev- enue. The situation here bears similarity to a state agency that in the process of contracting a certain number of plows before the snow season begins when it is not known beforehand how intense or mild next winter will be. Therefore, a large number of plows may be contracted needlessly or vice versa. The question becomes one of what is the most economical number of trucks to be contracted and that to be outsourced subject to a specified level of service. Developing a model that captures the randomness in the weather patterns, reduces the cost, and gives better performance aspects of contracting is our intent in this work. It is noteworthy to mention that the application of the Newsvendor problem has not been con- sidered in the arena of snowplowing activities. This is one of this paper’ s contributions. 1 Professor, Mechanical and Industrial Engineering, New Jersey Institute of Technology, Newark, NJ 07102. 2 Visiting Professor, School of Automobile, Chang’an Univ., China; and Professor, Civil and Environmental Engineering, New Jersey Institute of Technology, Newark, NJ 07102 (corresponding author). E-mail: chien@ adm.njit.edu 3 Professor, Civil and Environmental Engineering, New Jersey Institute of Technology, Newark, NJ 07102. 4 Research Assistant, Interdisciplinary Program in Transportation, New Jersey Institute of Technology, Newark, NJ 07102. Note. This manuscript was submitted on January 9, 2013; approved on November 13, 2013; published online on February 3, 2014. Discussion per- iod open until July 3, 2014; separate discussions must be submitted for individual papers. This paper is part of the Journal of Infrastructure Sys- tems, © ASCE, ISSN 1076-0342/04014006(7)/$25.00. © ASCE 04014006-1 J. Infrastruct. Syst. J. Infrastruct. Syst. 2014.20. Downloaded from ascelibrary.org by Nj Inst Of Technology on 04/13/15. Copyright ASCE. For personal use only; all rights reserved.