Stand on a busy big-city street corner at lunch time and you will witness a chaotic scene. Thousands of people are walking every which way, getting on and off buses, descending to subways, riding in cars, and walking in and out of buildings. Where did all these people come from? And where are they going? Until now, such questions were unanswerable–mysteries of the city. But no more.
Today, thanks to deep analytics, we can for the first time understand the complexities of cities in motion.
IBM Researchers have developed analytics software that provides accurate and meaningful information about massive numbers of peoples’ movements. These insights can be used by city managers to plan new transit routes, improve the efficiency of current transit systems, and coordinate the various transportation modes with a goal of making moving around in cities a lot more convenient and comfortable. The project, Insights in Motion, is a so-called First-of-a-Kind (FOAK) collaboration with transportation officials in Dubuque, Iowa, and Istanbul, Turkey.
A paper about Dubuque’s piece of the project, Dubuque Smart Travel, was presented Jan. 16 at the annual meeting of the Transportation Research Board.
Until now, city officials based their knowledge of transportation activities on a wide variety of information–including everything from trip and ticket sales data, to surveys of transit passengers, to actual counts of people on a particular bus or subway car at a particular time in a particular place. The problem is, all of these types of information only provide fragments of the bigger picture. A survey of people on a particular bus route, for instance, only tells you about the people who are riding the bus–not about people who are moving in the same direction at the same time via other means who might ride the bus under other circumstances.
The Insights in Motion technology draws on transit data, geo-spacial information, census records, points-of-interest information and data from cell phones and smart phones. The telephone data is completely anonymous so no individual’s privacy is compromised. By tracking the movements of thousands of people from place to place and correlating it with time and the speed of travel, the system understands the mode of transportation people are using and knows where they’re traveling to and from–whether its home, work, school or shopping. For city planners, it’s a revelation. “It’s like a blind person for the first time opening their eyes and seeing,” says Milind Naphade, leader of the IBM Research project.
The project got its start because IBM Research scientists were looking for a real city in which to test their Smarter Planet technology ideas. The small but aspiring Mississippi River town of Dubuque agreed to become a living laboratory for experiments. One of the goals there was to increase the use of the public bus system, so Naphade and his colleagues proposed developing a system for analyzing the transportation needs of a small city. They applied within IBM Research to make Dubuque Transit a FOAK program, jointly funded by IBM and a client. They were asked to also choose another city, a larger one in an emerging market, and ultimately, they picked Istanbul.
So they had two living laboratories that are as different as you can imagine: a small city with just 23,000 households and only one source of public transit, buses; and a sprawling city with 14 million residents, 1 million tourists annually, and a wide range of transit options–subway, light rail, express buses, traditional buses, minibuses and ferries.
They started in Dubuque. In addition to gathering mobile phone location data (with no personal information included) from telecom carriers, they also recruited 1,000 volunteers who own smart phones. The high-end handhelds are equipped with GPS and accelerometers, which makes it possible to track the volunteers as they travel around the city with pinpoint location and speed-of-movement accuracy.
Their task was challenging. They had to build a system that could, in a sense,”understand” what a city is, what people are, what people do, and how they move from place to place. They needed a digital model of a city that they could experiment with. They built a framework of knowledge about each individual’s activities, including detecting meaningful locations, segmenting trips, calculating the duration of stay, identifying the purpose of each trip, plotting origins and destinations of each trip, tracking the time of day, and specifying the mode of transportation. They even estimated the carbon footprint of each trip. Then they aggregated all of the information about individuals to generate city-wide statistics.
The team used the results to optimize the routes and schedules for the city’s bus transit system. The objective was to minimize the sum of operator costs, user costs and unsatisfied demand costs for the entire network. Costs aren’t just expressed in dollars. For users of the system, for instance, they include factors like waiting, walking and driving a car. Dubuque is now trying the information out on two pilot projects–new bus routes it didn’t operate before. “Having the data is crucial. You don’t have to operate on tea leaves,” says Chandra Ravada, director of the transportation department for Iowa’s East Central Intergovernmental Association. “You know where people are going. You can change things based on that. You’re designing your system not on somebody’s opinion but based on facts.”
Naphade and his team took the lessons learned in Dubuque and applied them in Istanbul. Their task there was way more complex. There were so many more people and modes of transportation. Also, in Istanbul, they didn’t track volunteers with Smart Phones. They had to rely on less precise records from regular mobile phones. But the technologies they had used in Dubuque also worked in Istanbul. The city’s transit authority is using the Insights in Motion tool to help design feeder bus routes connecting to the cities new subway lines. The goal is to reduce operating expenses by 40%, meet 37% more demand, reduce average commuter time by 60% and reduce per-traveler combustion emissions by 40%.
At the highest level, Naphade believes that Insights in Motion has tremendous potential to alter the relationships of cities and their citizens. “One thing I think about is how we have become slaves of the infrastructure rather than having the infrastructure work for us,” he says. “Cities should help people live their lives, not get in the way.”