I was standing on the platform at Manchester Piccadilly, staring at the departure board with that familiar mix of resignation and frustration. My train home to Sheffield was delayed again — the third time that month. The automated announcement apologised for “unforeseen circumstances,” and I remember thinking, “Surely someone can foresee these by now. Don’t they have data for this?” I wasn’t a data scientist. I was just an annoyed student who wanted to get home before midnight. But that question — about how data could predict and prevent these everyday frustrations — lodged itself in my brain and refused to leave.
Over the following weeks, I started paying attention to the data trails I left behind every day. My contactless card knew exactly where I’d been; my phone tracked my steps without me asking; the weather app somehow always knew when it would rain on my walk to the library. I began to wonder: if all this information is already out there, why aren’t we using it better? Why can’t we predict train delays, or hospital waiting times, or even which students might drop out before they reach their final year? The more I looked, the more I realised that data science and analytics weren’t just corporate buzzwords. They were tools for solving real, tangible problems — the kind that affect ordinary people like me, standing on cold railway platforms in the middle of winter.
When the time came to settle on a dissertation topic, I knew I wanted to use data to tackle a practical question rooted in the UK context. But I needed to move from vague frustration to a focused, researchable idea. I started by exploring what other students had already done. I found a collection of data science & analytics research topics for students UK 2026 and spent a couple of evenings working through it. Some projects analysed NHS waiting time data to identify bottlenecks, others used machine learning to model air pollution in London boroughs, and a few examined how social media sentiment correlates with consumer confidence in the UK economy. That range gave me the confidence that my own curiosity — around using publicly available transport data to predict service disruptions — could become a legitimate, researchable project.
With my direction clearer, I began to shape my study. I focused on the TransPennine Express route (the very one I’d been cursing that night), collected historical delay data, weather records, and passenger volumes, and started building a simple predictive model using Python. My supervisor helped me ground the work in the growing field of urban analytics, and the project started to feel less like a requirement and more like a genuine contribution — something that could, in a small way, help transport planners make better decisions and make life a little less miserable for the thousands of commuters who rely on those trains every day.
If you’re a student in the UK who wants to explore data science but doesn’t know where to start, look at the problems you encounter in your daily life. The delayed train, the overcrowded GP surgery, the air quality alert on your phone — these aren’t just annoyances; they’re potential research questions hiding in plain sight. Then explore what other students have already investigated, and use their work as a springboard for your own project. Data science isn’t about being a maths prodigy; it’s about asking good questions and being willing to learn the tools to answer them. You might just find that your best idea was waiting for you on a cold railway platform all along.