Master thesis; Deep Reinforcement Learning for Resource Scheduling in Railway Operations with Explainable AI

We are looking for a dedicated master’s student to join us in the Systems Engineering Unit at RISE.

The Systems Engineering Unit is part of RISE Mobility and Systems, focusing on Resource Optimization and System of Systems. The unit conducts projects together with industry and academia nationally and internationally. The research focus for the proposed thesis topics is within Resource Optimization and Decision support.

Background
Transportation systems need continuous development to meet future demands for sustainable transport. Expectations on these systems are constantly increasing regarding reliability, predictability, transparency, and efficiency. The railway sector has many strengths but also faces significant challenges in resource planning. Real-time traffic management and dispatch must account for delays, maintenance windows, and the limited availability of critical operational resources including rolling stock, crews, depots and network capacity.

Reinforcement Learning (RL) is an emerging AI method where agents learn to make decisions through interaction with their environment. In this project, you will explore how Deep RL techniques can be applied to railway engine scheduling in a dynamic, constraint-based setting. In particular, the work will also focus on Explainable AI (XAI) enabling stakeholders to better understand and trust AI-based decisions.

• Description
• We are offering a master’s thesis aimed at enhancing railway traffic management and resource allocation using AI.

The purpose of this thesis is to explore alternative DRL algorithms and evaluate how well they perform on the same planning task using existing custom-built simulation environment. The project also aims to investigate how Explainable AI (XAI) methods can help visualize and interpret the agent’s decisions making them more transparent and trustworthy to domain users.

Qualifications
We expect you to have a strong and solid knowledge of AI and machine learning, good programming skills in Python and an interest in solving complex problems. You are interested in deep reinforcement learning (DRL experience is helpful but not required) and in making AI systems more interpretable and user-friendly. Experience with data visualization or simulation is a plus.

Collaboration
You will be supervised by experienced AI researchers at RISE and work in collaboration with stakeholders in the Swedish railway sector. The thesis offers a unique opportunity to contribute to real-world planning challenges using cutting-edge AI methods.

Terms
Start Time: Preferable January or February

Scope: 30 hp.

Compensation: For an approved thesis project worth 30 credits, we pay a compensation of 39,990 SEK if one student and 30,000 SEK per student if more than one student.

Location: RISE Mobility and Systems, Västerås or Stockholm. Option to partially work remotely.

Welcome with your application!
We welcome applications from students passionate about making a significant impact on the future of transportation systems. Please specify in your application which thesis opportunity you are interested in.

To know more, please contact Zohreh Ranjbar zohreh.ranjbar@ri.se, 070 627 85 98. Applications should include a CV, recent transcript of records, and a code excerpt (example of a code file written by you, or your GitHub repository link). Candidates are encouraged to send in their application as soon as possible but at the latest by the 1st of December. Suitable applicants will be interviewed as soon as applications are received.

Join us in shaping the future of railway resource planning through innovation and cutting-edge research!