Hi I’m Bea, and I’m a postdoc in the Centre for Advanced Spatial Analysis (CASA) at UCL. I’m part of the AI4CI Smart City Design project, working on how we can use AI to improve our collective understanding of the London planning system.
My academic background is somewhat mixed. I started off doing a maths degree at the University of Warwick, and then I did my PhD in computer science at UCL, where I looked at computational modelling of neurodegeneration. Basically, at each step I’ve shifted into more applied work, which suits me well, as I’m really interested in how we can use the latest ideas to improve our understanding of pressing problems.
When I’m not working I like to be making things or arguing about politics at the pub.
What are you working on?
I’m currently working on a project with Adam Dennett and Adham Enaya about how people object to proposals for new residential developments in their area. Specifically, we’re extracting and collating datasets of these comments, and then using natural language processing to start identifying the patterns inherent within them. The goal is to understand why people object to plans, and how this varies by location and building type.
What excites you most about your field of research?
I think the potential to make a meaningful contribution is one of the most exciting things about the field. In our current work, we’re exploring issues connected to the UK’s housing crisis. The government has responded to the crisis by reforming the planning system and encouraging local councils to meet ambitious housebuilding targets. However, I think this doesn’t necessarily address the root cause of the crisis. Our research focuses on analysing what types of homes are being proposed and how local communities perceive them, using a rigorous, data-driven approach. The aim is to better understand where government policies are succeeding — and where they may be falling short, with the hope of informing better solutions.
Who has influenced your academic journey the most?
Overall, I think my academic journey has been more influenced by ideas, than individuals. I think science, and scientific understanding has a really important role to play in society – this belief has guided me down a more applied path. How we communicate our ideas, how we disseminate them, how we teach them – this influences my academic choices. Having said that I’ve definitely been lucky to know some really supportive academics across the departments where I’ve studied. I’m always impressed by those who not only conduct fascinating research but still make the time to engage with students and ECRs.
Can you recommend a paper you think we should read?
I recently read a paper called ‘Evaluating the World Model Implicit in a Generative Model’ which I found really compelling. In it they train a Large Language Model to learn a map based on data collected from cab rides in New York. What they find is that despite a surface level coherence (the trained model can recommend a route through the city) the resultant map is completely incoherent. I think this is so interesting, because it highlights the limitations of this type of AI model, and it’s limited on a task that humans just implicitly know how to do. Also the maps they produced are really cool – and a great example of data visualisation.
What are your ‘Top Tips’ for early career researchers?
I think it’s quite easy to become siloed in academia, but being at this early stage offers a unique chance to explore not just different ideas, but also different facets of academic work. So, I think what I’d recommend to other ECRs is to just get stuck in and experiment with new things. Whether it’s research, outreach, or teaching. Engaging broadly isn’t just personally enriching; it’s also a chance to help shape the research culture. By prioritising collaboration, equity, diversity, and inclusion, we can contribute to creating a more supportive academic environment that genuinely reflects our values.