Photo: Hindustan Times
Photo: Hindustan Times

Indo-Dutch project plans to tap cellphone data to decongest roads

The project will begin in June and aims to develop algorithms that will be able to predict traffic movement

New Delhi: When the Delhi-Gurgaon Expressway shut some of its lanes on a scorching summer day last week without warning, thousands of cars, sports utility vehicles and buses clogged a 15-km stretch in a traffic jam that took almost eight hours to clear.

For commuters, who crawled along the expressway at 1 km/h under a blazing sun that day, and many million more across Indian cities where traffic snarls have become a daily routine, an Indo-Dutch project to decongest roads by tapping drivers’ cellphone data could turn out to be a boon.

The project will begin in June and aims to develop algorithms that will be able to predict vehicular movement, said Ashish Verma, an assistant professor of transportation engineering at Indian Institute of Science (IISc) in Bengaluru.

If successful, the project will be able to harness data from mobile phones to reduce congestion on roads, particularly during peak traffic in the morning and evening. Traffic jams leave commuters frustrated and have resulted in increasing incidence of road rage.

Poor public transport, bad roads, a growing urban population and surging car ownership have combined to clog up Indian roads, in turn causing fuel wastage, loss in productivity and higher pollution levels.

“Everyone carries a cellphone when they are travelling. So why not use this data to capture traffic-related parameters which can be useful to develop various applications," said Verma.

Verma is leading the project from India with his counterpart Peter Sloot at the University of Amsterdam.

Other institutes working on the project include the New Delhi-based CSIR-Centre for Road Research Institute (CRRI) and Bengaluru-based M.S.Ramaiah Institute of Management; on the Netherlands side, academic partners include University of Amsterdam and Delft University of Technology.

Researchers said the project may lead to more efficient (and automated) traffic lights, and better information for public transport so that bus services are more efficient and reliable. The scientists plan to assess the kind of supercomputing power that will be needed to analyse the expected huge volume of data.

“In theory, this is a great assignment and concept which has the potential for tangible benefits. But it seems to lack commercial and financial viability unless more inclusive measures are taken for greater industry involvement," said Sanchit Vir Gogia, chief analyst and chief executive, Greyhound Research.

“Another concern is how security will be managed as the quantum of data is huge and any misinterpretation of data can cause even physical damage. The security aspect has to be talked about very seriously," he added.

To be sure, mobile phone information for this project can only be used if it conforms with the regulatory requirements in the country and if the data is kept anonymous, said Verma.

“We are not interested in who is moving where," said Verma, who is also president of Transportation Research Group of India, made up of transport professionals.

The research team will look into signalling data to find the location of a commuter at various time intervals. This would help them understand the person’s movement in the city within the transport network.

Signalling data can be provided by cellphone companies which locate the devices using the method of triangulation.

“We are developing an algorithm to trace vehicles and to understand how vehicles are placed on the roads using different mobile networks," said S. Velmurugan, principal scientist, traffic engineering and safety at CRRI.

“I would be interested to know whether the cellphone user is in a bus, a car, on a two-wheeler or is walking, depending on the speed and clustering."

The scientists will work on how to use the cellphone data and convert that into meaningful information for traffic workers. “It is a challenging exercise and quite interesting. A single source of data may not give me all information, so we may need other kinds of information sources, such as vehicles with GPS fitted in them or cameras installed by traffic police at various junctions," said Verma. “Once you have this basic data, live or otherwise, you can build various applications around this."

The memorandum of understanding (MoU) for this project was signed between the Netherlands’ ministry of infrastructure and environment and India’s ministry of urban development in 2013. The Netherlands Science Foundation is funding the project with €500,000.

The project also has industrial partners such as telecom company Vodafone Plc, the consultancy KPMG Llp and Mezuro BV, a Dutch company that looks at vehicular movements in a specific area on basis of anonymous data from mobile communications networks and cameras.

Big data analysis of traffic information from multiple sources is being used successfully in many developed countries. In the 1980s, Los Angeles used loop detectors or wires placed below the pavement to monitor traffic flow and synchronize traffic lights.

In 2013, International Business Machines Corp. (IBM) and the Dublin City Council agreed to use big data from multiple sources such as GPS-enabled cars, closed circuit television (CCTV) cameras and loop detectors to reduce traffic congestion in the Irish capital.

Close