Less Delays Through the Smart Use of AI: How the RAIL Lab is Transforming Train Scheduling
Unpredictable factors like failing technology, overactive badgers, erratic weather, and sick colleagues make train planning a highly time-consuming task. Currently, creating a schedule for complex railway hubs takes the Dutch Railways (NS) and ProRail two days of work.
In this recently published feature by NWO’s Onderzoek Magazine, the spotlight is on the groundbreaking work happening at the RAIL Lab. This collaborative initiative bridges the gap between academic research and industry practice to tackle unpredictable real-world disruptions. Through the lab’s efforts, researchers and industry partners have already optimized a complex, two-day train scheduling process down to just ten hours, with the ultimate goal of reaching a single hour.
A Strong Partnership: Industry and Academia
The RAIL Lab brings together researchers and industry partners to tackle real-world disruptions and create a more resilient railway network. Established in 2022, the lab is part of ICAI’s network and the long-term ROBUST program, which is partly funded by the Knowledge and Innovation Covenant (KIC) of the NWO.
This collaboration involves:
-
Industry partners such as Dutch Railways (NS) and rail infrastructure manager ProRail.
-
Academic institutions such as Utrecht University (UU) and Delft University of Technology (TU Delft).
Currently, five PhD candidates are working full-time on this “hub problem” alongside NS and ProRail.
Simulating the Perfect Schedule
One of the researchers, Issa Hanou, uses simulations powered by algorithms developed by her team to predict the best decisions to make when train traffic is disrupted. These smart algorithms can determine:
-
Whether a train should depart a station earlier or later.
-
Whether a train should run slower or faster.
While Hanou notes that it is unrealistic to expect a single, ready-made product to solve everything immediately, the progress has been tremendous.
“It is not realistic to think that researchers will come with a ready-made product. NS understands this well.” – Issa Hanou, Researcher
The Value of Collaboration
This initiative perfectly highlights the mutual benefits of academic-industry partnerships:
-
Fewer delays – smarter algorithms translate directly to fewer delays and a more reliable public transport system for daily commuters.
-
Scalable solutions – access to cutting-edge academic research fuels practical, scalable solutions for highly complex logistical challenges.
-
Tangible scientific impact – real-world data and practical insights enrich scientific inquiry, ensuring that research has a direct, practical application.
Bob Huisman, Lead Consultant of IT Innovation at NS, emphasizes that solving these massive societal challenges requires long-term commitment, shared funding, and a deep mutual understanding:
“Without this collaboration, we would never have come this far.” – Bob Huisman, NS
The full NWO article is available for download here.
