Investigating pedestrian behaviour in urban environments: A Wi-Fi tracking and machine learning approach

•Utilising merely objective movement information overlooks key walking behaviours.•Unsupervised machine-learning frameworks revealed four pedestrian behaviours.•Activity type and area knowledge represent the key behaviours in a high-street.•Increased temporal and spatial data granularity are needed in the tracing process.•Practical application of ML approaches requires standardisation of the process.