Keiran Suchak


I am a PhD Student in the School of Geography, based at the Leeds Institute for Data Analytics. Having studied a wide variety of subjects over the course of my education, I have developed an interest in the modelling and analysis of social systems. My PhD therefore focuses on the development of new methods for the simulation of pedestrian motion in urban settings at close to real-time, and is partnered with Leeds CIty Council. As part of my PhD, I have undertaken an internship with Leeds City Council exploring patterns in footfall at Kirkgate Market.

In a professional capacity, I have worked as a developer and analyst focussing on media analytics, the development of data processing pipelines and charting the provision of television channels across the globe. I have also recently taken up a 3 month Data Science Fellowship seeking to model all-age social isolation in Buckinghamshire with the Connected Places Catapult.

Academic History:

  • CDT Data Analytics & Society – University of Leeds (2017 – present)
  • MSc Mathematics – University of Leeds (2015 – 2016)
  • MSc Aerospace Dynamics – Cranfield Univerity (2013 – 2014)
  • BSc Physics – Imperial College London (2010 – 2013)

Professional History:



Research interests

My research interest focus on the application of modelling techniques to social phenomena – this has included the exploitation of graph automorphisms to aid with the modelling of how infections spread across networks, the use of regression models to estimate footfall at Kirkgate Market, and the implementation of geographically weighted regression models to explore the spatial variation in factors contributing to social isolation.

At present, my research focuses on the development of new methods for the real-time simulation of pedestrian system. The modelling of pedestrian systems is often undertaken through the use of agent-based models. Such models consider individuals in the system as autonomous agents which interact with each other and the environment around them, often resulting in the emergence of structure at a macroscopic level such as crowding or lane formation. These models are typically calibrated using historical data before being run, but often do not make use of up-to-date observations as they the proceed to model the system state. One of the reasons for this is there does not exist an agreed-upon way in which to incorporate new observations as the model runs. This project, therefore, aims to adapt data assimilation methods from the field of numerical weather prediction to combine the model state with the observed state with a view to improving the accuracy with which an agent-based model can simulate a pedestrian system.

The project ties in with an ongoing body of work which falls under the DUST project.


  • MSc Mathematics - University of Leeds (2016)
  • MSc Aerospace Dynamics - Cranfield University (2014)
  • BSc Physics - Imperial College London (2013)