ITU master’s thesis can improve train operation
Matthias Als and Mathias Bejlegaard Madsen have finished their MSc in Computer Science at ITU, and their master’s thesis has recently been published in a research journal. They have developed energy efficient timetables for trains that can reduce trains’ energy consumption, operating costs, and CO2 emissions without extending the passenger travel time.
It began with a beer in Scroll Bar at ITU during their first semester. Since then, Matthias Als and Mathias Bejlegaard Madsen have developed a strong friendship and achieved great academic results. They hold an MSc in Computer Science from ITU, and their master’s thesis was a successful culmination after five years of studying. The thesis has recently been published as a research article in Journal of Rail Transport Planning & Management.
They have developed energy efficient timetables for trains that can reduce trains’ energy consumption and operating costs without extending the passenger travel time. Furthermore, energy efficient timetables can contribute to the green transition.
“Replacing the current train machinery with more climate friendly alternatives can take many years. While waiting, energy efficient timetables can reduce energy consumption, minimize the CO2 emission, and contribute to the green transition right now” explains Mathias Bejlegaard Madsen.
Energy efficient timetables based on historical data from train operation
According to Matthias Als and Mathias Bejlegaard Madsen, energy efficient timetables are essential in order to use trains’ full energy potential.
“The energy consumption increases faster than the train driver increases the speed. Consequently, you can save a lot of energy by driving slower. However slow driving is not a solution, because it prolongs the passenger travel time” says Matthias Als. Therefore, the two computer scientists have focused on two parameters: the trains’ energy consumption and how the energy consumption influences the passenger travel time.
A key feature in Matthias Als and Mathias Bejlegaard Madsen’s thesis is that they model actual energy consumption based on historical data from train operation. This method includes taking the train driver’s behavior into account, and it makes it possible to predict the energy consumption on a specific timetable. When the energy efficient timetables are assessed based on their energy consumption and the driver’s behavior, it produces a more realistic result.
Every timetable needs optimization based on the timetable’s energy consumption and passenger travel time. Matthias Als and Mathias Bejlegaard Madsen have developed a genetic algorithm to model the timetables. It pairs and mutates the timetables and thereby creates the most optimal energy efficient timetables.
Matthias Als and Mathias Bejlegaard Madsen apply their method on a real-world case from a large North European train operating company. The considered network consists of 107 stations and junctions and 18 periodic timetables for nine train lines. Their results show that for an entire network, a reduction up to 3.3% in energy consumption and 4.64% in passenger travel time can be achieved.
Matthias Als and Mathias Bejlegaard Madsens wrote their master’s thesis in collaboration with Cubris and under guidance from Rune Møller Jensen, Assistant Professor at the Computer Science Department at ITU.
From ITU to the industry
Even though Scroll Bar, assignments, and exams at ITU is a closed chapter, they look back at their time of study as a period that prepared them for their careers.
“We really enjoyed programming when we studied, and we wanted to do that all the time during our education. Today I am happy that we also learned about design perspectives, business needs, and user orientation. We have obtained many different competencies to approach tasks and solve the challenges we meet in our jobs” explains Matthias Als. He is working as Senior Data Scientist at Ecco in the department of Data & AI. Mathias Bejlegaard Madsen is working at Cubris where he is Team Lead in the department of Optimizing & Data Science.
Ditte Ørsted Johansen, Press Officer, phone +45 25 55 04 47, email email@example.com