CASA-UCL will show a set of visualisations of spatio-temporal datasets revealing urban dynamics and citizens' activity and mobility patterns. Some examples are shown below:
Pulse of the city. Using three weeks worth of Oyster Data, travel time information and shortest paths between stations we can calculate the time and intensity of loadings between every station in the Underground system. Between Euston and Warren Street there may be nearly 6.500 people around 8.30 a.m.
Geolocated Tweets and estimated activity patterns. A cross-cluster analysis of tweets against activity points shows how urban activities vary with time and space. The analysis results in a probability-based one-to-many link-map between Twitter instances (tweets) and activities. Results are displayed in the following animation of Geolocated Tweets (Tweeted weekdays in Central London). Colours reflect estimated activity (Red: Offices, Yellow: Retail, Green: Leisure, Cyan: Rail). Collection dates: 10/2013-05/2014, Source: Twitter API.
Simulating pedestrian route choice behaviour under transient traffic conditions. The following animation is displaying outputs from a model of dynamic assignment of pedestrian routes responsive to pedestrian traffic conditions (crowding). The cost of each routing option is determined using agent-based feedback and choices are made over continuous space abstracted into a conditional directional network.
The colour of each pedestrian reflects deviation from his preferred walking speed (red: high deviation, green: low deviation).