Proceedings of the 2011 Winter Simulation Conference
S. Jain, R.R. Creasey, J. Himmelspach, K.P. White, and M. Fu, eds.
STRATEGIC CREW PLANNING TOOL IN RAILROAD: A DISCRETE EVENT SIMULATION
Kiran Chahar
Clark Cheng
Yudi Pranoto
Norfolk Southern Corporation
1200 Peachtree St. NE
Atlanta, GA 30309, USA
ABSTRACT
Norfolk Southern (NS) has developed a strategic crew planning tool to evaluate the impacts of crew rules
changes and train service changes on crew utilization and train on-time performance. This tool has three
major components, a discrete event simulator, a crew deadheading engine, and a crew pool size analyzer.
A Flash based animation and reporting user interface helps users identify bottlenecks in specific areas of
the rail network. This tool is integrated into a suite of planning tools used in NS. The impact of crew mark
off rates on train performance is discussed in a case study.
1 INTRODUCTION
Freight trains are generally operated by an engineer and a conductor. The engineers and the conductors
belong to specific crew pools depending on the territories and on the types of crew pools. Every crew
pool has at least one home terminal and most crew pools have one or many away from home terminals.
Figure 1 below represents the crew operation between terminals. Home terminal means that the crew has
a home in that city and there is no lodging cost at Home. Away-from-home terminal implies that the crew
needs a lodging arrangement for resting in that city. Home and away-from-home terminals have different
business rules, for example, a crew waiting excessively at an away from home terminal can claim for de-
tention payment while a crew at home cannot claim for detention. Each crew pool has a set number of
jobs, called turns. Each turn is filled by one crew through a job bidding process where qualified crew with
highest seniority level will get the turn. The turns form a queue on a crew board, generally made up of
turns of the same pool to wait for their assignment to the trains. When the train arrives at its destination,
the crew will tie up at another crew board according to the crew tie up rules. After at least 10 hours of
rest, the crew will be available for his/her next assignment. Above mentioned freight train operations are
modeled in our strategic crew planning tool.
Optimization and simulation techniques have been utilized to model train crew assignment problem.
Crew assignment problem in European railroad has been mathematically modeled by splitting this prob-
lem into two sub-problems of crew scheduling and crew rostering (Caprara et. al 1997). Freight train crew
HOME AWAY-FROM-
HOME
Terminal A
Terminal B
Figure 1: Representation of the crew operation
3693 978-1-4577-2109-0/11/$26.00 ©2011 IEEE