Return to Projects List

[NTC2016-MU-R-05] Methods for Improving the Reliability of Transportation Systems

Paul Schonfeld, George List
University of Maryland, North Carolina State University
Subject Area: 
Collaborative Research


Reliability in transportation systems is a measure of predictability and consistency in departure and arrival times. It may be measured as the probability of arriving (or possibly departing) within a specified time span (or “window”), the variance of arrival, departure or service times, or the generalized cost of the variability. Reliability is one of the key aspects of service quality and system performance to be considered in planning, designing and operating transportation systems. The reliability of transportation systems may be improved in various ways involving overall system design, the detailed design of facilities and vehicles, maintenance of facilities and equipment, routing, scheduling, traffic management, terminal operations, control of vehicle movements, provisions for reserves or slack in various system components, and preparations for contingencies. While seeking to develop general methods applicable to various kinds of transportation systems, the objective of the proposed study is to improve freight transportation reliability in road networks. Furthermore, this study will focus on three important aspects of transportation system reliability, namely (1) the development and maintenance of reliable networks, (2) real-time vehicle dispatching decisions, and (3) resulting resource requirements, especially fleet sizes.

In analyzing the development and maintenance of transportation networks, we will develop methods for evaluating candidate projects or alternatives based on their reliability effects, in addition to various effectiveness measures, including infrastructure costs, user costs and benefits, environmental impacts and some external economic impacts. The two major improvements over conventional methods for transportation investment planning and scheduling will be (a) their explicit consideration of reliability measures jointly with other effectiveness measures in the evaluation and optimization processes and (b) their consideration of quantifiable interrelations among alternatives. Interrelated alternatives are those whose benefits and/or costs depend on which other alternatives are implemented at what times. Consideration of interrelations is very important for transportation networks because changes in network components shift traffic and thus affect the benefits of improvements to other components. Temporary interruptions in resource availability for maintenance purposes may also shift traffic and their effects depend crucially on whether the closed elements are in parallel or in series. The costs, budgets and other resources available for various alternatives may also be interrelated. The currently available analysis methods are relatively well suited for dealing with mutually exclusive alternatives or independent alternatives but quite inadequate for dealing with realistic numbers and complexities of interrelated alternatives. By extending methods already developed by our team (Tao &Schonfeld 2005, 2006, 2007, Wang & Schonfeld 2005, 2008, 2012, Shayanfar & Schonfeld 2015, Yang et al 2015) we expect to develop methods that not only evaluate interrelated alternatives appropriately, but also optimize the selection, sequencing and scheduling of those alternatives, subject to constraints on reliability, continuity, budgets, various resources, implementation times, fairness and other factors.