Xiaohongshu
Reflections on the 2026 Sustainable Housing Hackathon
Author: Fiona Pacolli, BA Geography (International) with Social Data Science
Recently, I had the opportunity to attend the Sustainable Housing Hackathon, organised by homefolk and Students’ Union UCL’s Social Enterprise Team. We were challenged to prototype an open-access mapping tool that allows users to input a London postcode and obtain detailed surface area data as an output. I joined a table on the day and quickly found myself working with teammates at different levels of study, from first-year undergraduates to post-graduates.
As a second-year student studying Geography with Social Data Science, I am especially interested in topics such as uneven urban development, spatial analysis, and housing provision. So, this hackathon was a great opportunity to apply my technical skills and learning from modules like Data Analysis, Urban Geography, Development Geography, and Geocomputation.
The event began with an introduction to homefolk and its decentralised co-housing model. homefolk (CIC) is a youth-led social enterprise pioneering a unique model for affordable, sustainable, and community-owned homes in paved urban spaces. Cities like London have significant variability in marginal paved spaces, yet there are no accessible tools to visualise this potential. With that in mind, my team and I set out to build a functional mapping tool using open-source spatial data to highlight where homefolk’s tiny ‘village’ model could be implemented.
As a team, we started with ideation, choosing open-source datasets and assigning roles based on our strengths. I focused on data wrangling in R, cleaning and filtering the ONS Postcode Directory to create a London-specific postcode-to-Output Area lookup table, which we then exported for Python use. This was a crucial step, as linking each postcode to its corresponding Output Area allowed us to aggregate and layer other spatial datasets accurately. Then, a teammate with stronger Python experience compiled OpenStreetMap buildings and traffic data as well as Ordnance Survey green space data (also filtered to London) into a GeoPackage and built the website application using the Streamlit library and AI.
Within a few hours, we had a functional prototype that mapped surface area data and displayed surface area-by-type percentages based on the postcode inputted. With just 30 minutes left, we put together a short presentation outlining our process, limitations, and suggestions for how homefolk could develop the tool further.
The event also featured a presentation about social enterprises, exploring what sets them apart from other business models. At its core, a social enterprise is a business with a social and/or environmental mission which reinvests or donates at least half of its profits to further that mission.
By walking us through the spectrum between traditional non-profits and traditional for-profits, I came away with a much clearer understanding of the financial and legal distinctions between social enterprises and non-profits, having previously underestimated how structurally different they are.
I am particularly proud of our achievement and team effort as this hackathon showed me that I can take what I am learning in my degree and directly apply it in collaborative, time-pressured environments to prototype, in this case, an open-access mapping tool that both the Social Enterprise Team and homefolk representative were impressed by. This experience left me feeling inspired, as I am finding myself more drawn to opportunities where I can address real issues and contribute to innovative and meaningful progress using my skill set and knowledge.
