Planning public transport in a big city like Singapore is a hugely challenging task. A new MRT station costs tens of millions of dollars, and a new line billions. So it’s important to build them where they will bring the greatest benefit. Knowing where people travel to and from, when they travel, and what form of transport they use is essential for effective planning.
And now, thanks to data science, important new insights are available for city planners. DataSpark – SingTel’s data analytics subsidiary – gathers the location data that users’ mobile phones automatically provide as they move through the SingTel network and analyses this information to gain insights into where they are going and what they are doing.
DataSpark data scientist James Decraene has been investigating the possibilities of using this data to improve public transport planning. He says that by analysing where people are at, during particular times, their home and work locations can be determined with a high degree of accuracy along with the routes taken and the mode of transport used when they travel.
All data used by DataSpark is encrypted, anonymised and aggregated to be compliant with SingTel’s data governance framework and with Singapore’s Personal Data Privacy Act.
Simulating transport scenarios
“The most interesting recent research we have been doing is simulation of human mobility for sustainable city planning,” he says. “We can approximate where people live and work, we know the location of the stations and we can assume that if people live close enough to an MRT station they will take the train to work. Such assumptions can be validated through simulation and using actual data. We can also determine the average times at which people go to work. So we can simulate and validate using actual data a typical flow of traffic given the existing transportation infrastructure and even evaluate the potential impact of new MRT lines and residential estates.”
The biggest challenge facing Singapore’s public transport planners is managing peak hour traffic and smoothing out those peaks. The Singapore Government has tried offering free trips from certain stations before 6.00am and is also offering employers financial incentives to encourage them to implement more flexible working hours.
But the results of the free trips scheme have been mixed, Decraene says. “If someone goes to work at 10.00am they are not going to use the free trip, and if they are wealthy they won’t worry about saving a few pence a day.”
Dealing with the peak hour problem
DataSpark’s access to mobile location data and its ability to analyse that data will, he says, enable it to predict the effectiveness of such peak hour management schemes. “We have the traveller profiles and the simulations within our model. So we can say, for example, that 10 per cent of the population will switch to off-peak travel. In fact, we can be much more precise and say that only one per cent of the wealthy will switch but 20 per cent of low income earners will.”
He’s convinced that what has been achieved to date is only a fraction of the analysis that will be possible when more location data becomes available and when, with experience, the analysis becomes more sophisticated.
Opening up new possibilities
Transportation planning means connecting a lot of different things and only by having a holistic view and putting all this information together can you implement the right new measures and build better transportation infrastructure.
“Through drawing a complete picture of Singapore where people are the central components of the city, we can best assist the complex task of city planners: for the first time, the lifestyle, lifecycle, quality of life and social aspirations of the people can be accounted for in a precise and comprehensive manner thanks to big data science. The end goal is to provide data-driven urban solutions where the potential for happiness of all residents, regardless of their background and social status, is maximised,” Decraene says.
“For instance, our simulation platform can effectively measure what will be the impact in terms of train crowdedness, accessibility to transportation and amenities given a new city plan where new residential estates, office parks, roads and MRT lines are designed. Future data science R&D will also focus on the impact on housing costs and the environment. This is especially critical for Singapore if the city is to reach a population of 7 million by 2030. The question is how can we ensure the quality of life of the residents is maintained and ultimately improved during this difficult transition.”
These insights will help to more effectively manage traffic and plan Singapore’s next generation of transport facilities, making Singapore a better place to live and work.
For more information on how data analytics can help make smarter decisions about transport planning, get in touch at email@example.com.