Intelligent sleeping car berth inventory control system in the case of variable capacity of railcars

• Author: Milica Šelmić,Dragana Macura,Dušan Teodorović
• Published date: 01 January 2019
• Date: 01 January 2019
• DOI: http://doi.org/10.1111/itor.12465

Intl. Trans. in Op. Res. 26 (2019) 270–288

DOI: 10.1111/itor.12465

INTERNATIONAL

TRANSACTIONS

IN OPERATIONAL

RESEARCH

Intelligent sleeping car berth inventory control system in the

case of variable capacity of railcars

Dragana Macuraa, Milica ˇ

Selmi´

cband Duˇ

san Teodorovi´

cb

aDepartment for Railway Engineering, Universityof Belgrade, Faculty of Transport and Traffic Engineering, Vojvode

Stepe, 305, Belgrade, Serbia

bDepartment for Operations Research in Traffic, Universityof Belgrade, Faculty of Transport and Traffic Engineering,

Belgrade, Serbia

E-mail: d.macura@sf.bg.ac.rs [Macura]; m.selmic@sf.bg.ac.rs [ˇ

Selmi´

c]; dusan@sf.bg.ac.rs [Teodorovi´

c]

Received 24 September 2016; receivedin revised form 14 August 2017; accepted 23 August 2017

Abstract

Managing ticket reservations in passenger rail transport is a complex task. A hybrid algorithm is developed

as a decision support tool. The aim of the proposed algorithm is to provide real-time decisions on seat

inventory allocationtaking into account relevant informationfrom the past. This paper considers the revenue

management application in passenger rail transport with variable capacity of sleeping cars. Uncertainty of

capacity availability is embedded in the developed algorithm througha payoff table,while an Artificial Neural

Network is used as a tool for making real-time decisions. The algorithm is tested on hypothetical data.

Keywords:revenue management; rail transport; variable capacity; artificial neural network; payoff table

1. Introduction

For many travelers, night trains are a very good alternative for overnight travel. Night trains

both offer seats in their day cars, and berths in their sleeping cars. For example, Cologne–Prague,

Zurich–Vienna, Amsterdam–Zurich, and Munich–Florence/Rome in Europe, or New Orleans–

San Antonio–Los Angeles, and Seattle–Portland–Los Angeles in the United States are some of

the popular night train routes. In order to improve matching sleeping car supply with sleeping car

demand, railway operators have introduced various differential pricing schemes. Fifty years ago a

number of airlines faced the problem of empty seats on some flights whereas some air carriers met

the problem of passengers exceeding the offered airline capacity. The differential pricing system

(holiday packages, special fare, discount for young travelers, etc.) was introduced as a result of an

effort of airlines to better match a demand for each flight with its capacity, and thus increase the

revenue, and a revenue management (RM) policy was born.

C

2017 The Authors.

International Transactionsin Operational Research C

2017 International Federation ofOperational Research Societies

Published by John Wiley & Sons Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main St, Malden, MA02148,

USA.

D. Macura et al. / Intl. Trans. in Op. Res.26 (2019) 270–288 271

The main objective of the RM concept is selling the right product/service to the right customer

at the right price at the right time. This concept can be applied when solving problems with the

following characteristics: a fixed amount of resources available for sale; the resources to sell are

prone to expiring (time limitation for selling the resources, after which they do not have any value);

different customers are willing to pay a different price for the same product/service. As we know, all

these factors are present in the transport industry. The main goal of RM is allocation of available

capacity so as to maximize profit.

RM is widely used in the transport industry, primarily in the airline sector. Railway and airline

industries differ in network structure, price differentiation, booking lead times, and market compe-

tition (Helve, 2015). These characteristics lead to different approaches to RM applications in these

two areas. The first scientific paper on RM application in railway was published in 1996 (Strasser,

1996).

Amtrak, an intercity passenger rail company in the United States, is a pioneer in the application

of RM and introduced the practice almost 25 years ago. Since then, the majority of other railway

companies have followed Amtark, such as SNCF (Soci´

et´

e Nationale des Chemins de fer Franc¸ais)

in France; GNER (Great North Eastern Railway) in Britain; DB (Deustche Bahn) in Germany;

VR Group in Finland. A number of these practices of RM concept implementation are described

in Helve (2015).

In many developing countries, railway operators cannot know the available rail capacity in

advance due to technical failures and repairs of rail cars. And this situation of availability of rail

capacity being unknown in advance is exactly what we consider in this paper. A nested reservation

system with variable capacity is analyzed. The main contribution of this paper is developing a

decision support system that enables relating real-time decisions to every new passenger’s request,

i.e. whether to accept it or not. The proposed algorithm is based on an Artificial Neural Network

(ANN). Uncertainty of rail car readiness is handled by a payoff table.

The paper is organized as follows. Section 2 introduces a brief, relevant literature review

of RM concept and its applications in the rail industry. After the literature review, Section 3

presents a mathematical formulation of the RM concept of variable capacity applied in the model.

Section 4 explains our hybrid algorithm based on an Artificial Neural Network and payoff table.

Results and discussion are summarized in Section 5. The last section consists of concluding remarks,

other possible applications of the model, and future research.

2. Literat ure review

The RM concept was first applied in air transportation. “Tourist Class”, “Business Class”, holiday

packages, special fares for young travelers, etc., were for the first time advertised and launched in

the 1970s by the Australian carrier “Qantas”. American Airlines introduced “Super Saver” fares in

1977 (McGill and van Ryzin, 1999). Research in the field of RM started in the late 1960s and early

1970s of the last century.Littlewood (1972) studied RM problem in the case of a single leg with two

fares. He suggested that discount fare bookings should be accepted as long as their revenue value

exceeded the expected revenue of potential full-fare bookings. This simple rule could be treated as

the beginning of the RM philosophy. Overbooking problems appear when some passengers with

valid reservation and validticket cannot get vacant seats on their planned flights. Pioneering papers

C

2017 The Authors.

International Transactionsin Operational Research C

2017 International Federation of OperationalResearch Societies