Cracking Enigmas with Alan Turing Institute’s Data Study Group

A blog by Ajaykumar Manivannan

In 2024, as a Ph.D. student in Canada, it was my dream to work at the Alan Turing Institute (ATI) in the U.K. For many, the name evokes the legendary scientist, whether you follow the trends in computing or simply enjoyed the brilliant 2014 movie "Imitation Game". Even as an international student, I recognized why establishing an institution like ATI was both inevitable and necessary—a premier national institute uniting talent from academia, industry, and government to advance Artificial Intelligence (AI) and Data Science.

Now, as a research fellow at the University of Leeds, I learned about ATI’s Data Study Group (DSG)—a week-long (occasionally two-week for online batches) collaborative hackathon bringing together 30 to 40 talented participants (M.Sc. and Ph.D. students, and postdocs) to address 3 to 4 challenges proposed by government, academia, or industry partners. Enthusiastically, I applied to the online edition of the January 2025 cohort. It presented an invaluable opportunity to tackle real-world problems with sensitive datasets that aren't always publicly accessible. The application was straightforward, comprising 3 to 4 questions requiring succinct answers reminiscent of creating impactful slogans.

Receiving my acceptance email filled me with a joy akin to joining the team tasked with cracking the Enigma machine. This cohort’s theme addresses pressing contemporary issues closely aligned with my postdoctoral goals: climate change adaptation strategies. After expressing my preferences among the available challenges, I was thrilled to join the team working on the ScotRail challenge. This initiative investigates how rail and road data can support sustainable mobility, reduce car miles, and enhance rail network resilience, directly contributing to Scotland’s government ambition to achieve Net Zero emissions by 2050.

Scotland’s railway system, like any transportation network, operates as a spatially networked system comprising nodes (stations) connected by physically embedded links (tracks). People’s mobility represents the dynamic flow within this network, analogous to blood flow in our circulatory systems or phone calls traversing telecommunication networks. Viewing these as interconnected systems enables us to assess the impact of events—such as station closures (e.g., the Ayr Station fire) or heightened demand (e.g., Glasgow/Edinburgh during the Edinburgh Fringe Festival)—on railway operations and road traffic. We had access to one year of rail operations and passenger movement data that is not publicly available, where we could see the flow of people at every instant in time through hundreds of stations spread across Scotland. Our analysis clearly demonstrated that disruptions propagate more robustly when the ScotRail network is considered a single interconnected system, influenced by events both nearby and distant. This project allowed me to showcase both the potential of Network Science and my expertise within this field. Similarly, my teammates contributed crucial and diverse skills in computing, carbon emissions, visualization, geographical data science, and software development, providing invaluable learning experiences.

The collaborative hackathon was intense. The limited timeframe isn't intended for comprehensive answers but to develop an overarching framework by effectively collaborating within an interdisciplinary team under time constraints. I learned when to step up, step down, step in, and step out as circumstances demanded. It's a true test of leadership and teamwork. Clear, sustained communication proved even more essential than subject expertise.

On a personal level, this experience was profoundly reflective. It validated my academic journey, demonstrating how my subject expertise and technical skills could contribute effectively to real-world solutions. It also showed me how much I have grown in collaborative environments since my undergraduate group projects. Indeed, collaborative teamwork is a skill that, like programming, must be cultivated intentionally. Soft skills enable the effective application of technical skills.

I highly recommend DSG to anyone looking to gain practical experience with complex datasets and tackle critical research questions of our time—whether you're from social sciences, law, political science, or an engineer in physical sciences with data science expertise. Your voice matters. You can contribute and shape the future of AI, irrespective of the distance your field of study might feel like from computing or AI. DSG is managed by a highly competent, friendly team of scientists, technical experts, and administrators; rest assured, you're in excellent hands. Most importantly, remember to enjoy and savor the experience.

I would like to extend my sincere gratitude to my supervisor, Viktoria Spaiser (UKRI Future Leaders Fellow), whose generous support, genuine interest in my personal development, and passion for climate change adaptation were instrumental in making this experience possible.

Ajaykumar Manivannan
Research Fellow
School of Politics and International Studies (POLIS)
University of Leeds
a.manivannan@leeds.ac.uk