Observing Cabs on a Regular Tuesday
As a result of participating in the prototyping weekend conducted during UPSingapore, I was able to massage and mangle a number of datasets, graciously made available by private and public organisations. One of the datasets that I managed to sink my teeth into was the collection of geocoded cab data collected over a 24 hour period on the 15 of May 2012 for 15,694 cabs.
From the data, I was hoping to observe the behavioural patterns of cabs in Singapore.
I started by comparing the differences between available and occupied cabs.
Based on this comparison, a couple of behaviours reveal themselves. As expected is the concentration of available cabs to the east, lined up at Changi Airport. Whilst the rest of the available cabs are generally distributed throughout the heartlands, possibly performing random walks in search of passengers. Occupied cabs, on the other hand, tend to favour major roads and expressways. Also to note is the lack of available cabs along the southeastern coast of the island.
Below is the animated visualisation, visualising cabs from 4pm to 10pm.
Central Business District
The dataset is pretty expansive, affording us to zoom into regions to take a look at how cabs behave at a more localised level.
Bounding ourselves to the Central Business District, it might be interesting to observe their behaviour as they cater to the peak-hour crowd, knocking off from work.
An alternative perspective to the data is to focus on the pick-up points or destinations. With this, the focus returns to the Central Business District, plotting all routes that either originate from the area, or end within the district.
To note is the behaviour of passengers, rather than cabbies, that passengers seem to prefer to drop-off on main roads, whilst the pick-up locations are more dispersed.
Fans of the Final Fantasy series might note the resemblance between the visualised routes with Lifestream of Final Fantasy 7. Unintentionally, I find this to be an apt analogy.