AI/maskininlärning för att uppnå strategiska mål inom järnvägssektorn
Information
Författare: Alma Bäckman, Kajsa LindforsBeräknat färdigt: 2021-06
Handledare: Malin Jakobsson, Johan Warenlöv
Handledares företag/institution: PA Consulting
Ämnesgranskare: Anders Arweström Jansson
Övrigt: -
Presentationer
Presentation av Alma BäckmanPresentationstid: 2021-06-02 13:15
Presentation av Kajsa Lindfors
Presentationstid: 2021-06-02 14:15
Opponenter: Carl Johan Casten Carlberg, Elsa Jerhamre
Abstract
A well-functioning and competitive transport system is of great importance for Sweden’s economic growth. The railway system has a central role in this matter. Furthermore, the Swedish railway system plays an important part in achieving international and national goals, such as Agenda 2030 and the Swedish transport policy goals. The railway industry is currently being digitized to a further extent. The railway is becoming increasingly connected and data is generated and collected continuously. This study aims to explore the potential of artificial intelligence (AI), with a focus on machine learning and optimization, on the Swedish railway system, to achieve strategic goals. The focus is mainly on sustainability and economic aspects. Potential challenges and how to overcome them are also studied. To achieve the purpose of the study a case study of Stockholm’s commuter rail has been used.
To explore the aim, a literature study together with interviews has been conducted. The literature study consists of a background explanation of the railway industry and Stockholm’s commuter rail, as well as how AI applications can be implemented. The interviews have been held with key players in the Stockholm commuter rail industry, along with experts and scientists. These have given valuable information regarding their thoughts on the potential of AI in the railway industry.
The result shows that AI has the potential to improve maintenance and traffic management and thereby create a more attractive and efficient operation of the Stockholm commuter rail industry. Further, this may lead to achieving strategic goals: international, national and commuter rail related goals. Due to the limitation of this thesis, it is difficult to draw conclusions based on the results regarding specific methods and algorithms that would be best suited for applications in the railway. Several challenges were identified for using AI to achieve strategic goals. To overcome these challenges, a combination of various actions is needed. Based on the results it is clarified that the industry is permeated of willpower, but that sufficient investments and initiatives are still largely lacking to overcome the challenges that exist in Stockholm’s commuter rail traffic.