visual explorations of urban mobility

Singapore's public transport system requires passengers to tap their smart card passes when boarding and alighting subways and buses. While these actions translate into different, distance-based, fare prices for travelers, they also offer perspective on the passengers aboard those vehicles at any given time.

This interactive application puts that knowledge literally at your fingertips. A multi-touch interface enables users to actively explore Singapore's bus network, and see where most passengers get on and off buses, how people connect between the island’s stations, and the way these patterns changes throughout the day.

Visualizing data on a large-scale multi-touch surface allows for the application of novel interaction techniques to engage a broad audience. The fluidity of the visualizations, with smooth transitions and highly responsive interactions, creates an enjoyable user experience to keep users engaged, and to stimulate an informed debate.

Users can switch between various visualization modes to get different perspectives on the same data set. The selected bus line, the highlighted bus stops, and the currently chosen time stay consistent in all modes, while the application seamlessly animates through the different visualizations.

The figure on the far left displays multiple bus service charts with glyphs showing the boarding and alighting of passengers at each station, in order to compare them over time. A user may also focus on a single, selected bus route on the map of Singapore, as shown in the near-left image.

The timeline of this visualization doubles as a histogram representing passenger load. In these two histograms of daily passengers, the temporal patterns of traffic peaks are visible. We found the morning peak is shorter in time, while the evening peak is spread out. We can intuit that this is due to a wider duration of time in which people leave the office, or are active at night such as running errands or having dinner before retiring for the day.

The user can quickly explore through time, altering the range dynamically from 30 minute intervals up to an entire day by adjusting a time slider, resulting in aggregations over the equivalent time interval.