A Combinatorial Approach for Railroad Operating Plan Design

By: Dr. Ravindra K. Ahuja


Abstract: 
A freight railroad's Operating Plan dictates the movement of railcars,
blocks, trains, crew and locomotives in a railroad.s network. It consists of
consolidating railcars into blocks of cars, determining train schedules (how
many trains to run; the origin, destination, and route of each train and the
days of operation; the train arrival and departure times for each station at
which it stops), and the assignment of blocks of cars to trains. The
railroad operating plan is similar to the schedule design problem
encountered in airline operations and is a very-very large-scale decision
problem containing billions of decision variables. We will describe a
decomposition-based approach to solve this decision problem through
combinatorial algorithms (construction methods, very large-scale
neighborhood (VLSN) search, and minimum cost flows) and present
computational results of these algorithms on real-life decision problems. We
will also share with the audience the lessons learnt in working with
industries and what works and does not work in practice.