Synthetic Modelling of Pedestrian Movement - Tallinn case study report

Abstract

This paper builds towards the argument that pedestrian traffic in the city can be successfully simulated with agent-based computational models if pedestrians' movement patterns are appropriately studied first. Furthermore, such simulation models, when finely calibrated and supported by onsite observations, allow planners to evaluate different urban design scenarios.We present a pilot study carried out in the centre of Tallinn, and discuss a way of how pedestrian movement simulations can be conceived. In the pilot study we recorded some 120 traces of pedestrians' movement and developed a prototype of an agent-based computational model to simulate this movement. Additionally we investigated the possibility of including solar analysis into the computational model. Already this short exercise offered us some interesting insights into how certain spatial qualities and features can drive pedestrian traffic making urban walkers to verge off the shortest routes. The pilot study was carried out in the context of the High Street project [1] for turning the centre of Tallinn into more pedestrian friendly area by redesigning urban space, calming vehicular traffic and creating new opportunities for businesses to flourish.