Trip plan generation using optimization: A benchmark of freight routing and scheduling policies within the carload service segment

The rail freight carload service segment enables the distribution of freight volumes down to the unit of single rail cars, and stand as an important alternative to road transportation. However, this service segment is often associated with significant uncertainties and variations in daily freight volumes. Such uncertainties are challenging to manage since operating plans generally are established long in advance of operations. Flexibility can instead be found in the way trip plans are generated. Previous research has shown that a commonly used trip plan generation policy does not exploit the available flexibility to the full extent. In this paper, we therefore suggest an optimization-based freight routing and scheduling (OFRS) policy to address the rail freight trip plan generation problem. This OFRS-policy generates trip plans for rail cars while still restricted by the customer commitments. The policy involves a MIP formulation with a continuous time representation and is solved by commercial software. We apply the OFRS-policy on a case built on real data provided by the Swedish rail freight operator, Green Cargo, and assess the performance of the policy comparing the current industry practice. The results show that by using the OFRS policy, we can achieve a reduction in the total transportation times, number of shunting activities and potentially also a reduction in the service frequency given the considered transport demand.

► This paper propose an optimization-based policy for trip plan generation.
► The policy is benchmarked against the current industry practice.
► We are able to show on significant improvements potentials.
► Weekly service frequency can be reduced with up to 40%, bringing major cost savings.
► Results are derived with respect to the currently valid customer agreements.