Japan seeks unheard-of new uses for cell location data

Big Brother by no means the only interested party

Japanese operator NTT Docomo plans to use its store of location data to work out where to build more houses, and how many people get stuck during an earthquake.

Those are two just of the potential applications for the huge amount of location data stored by mobile network operators, and where Docomo and the University of Tokyo will be starting their Mobile Spatial Statistics project with a view to finding other ways of exploiting the (anonymous) data for fun and profit.

The project site is entirely in Japanese, so we're grateful to Penn Olson for an explanation and comments from Docomo itself.

Mobile operators always know where their customers are, with a variable degree of accuracy dependent on the cell density. Most attention is on tracking individuals, but there's real value in knowing where the unnamed masses spend their time and it is that data which Docomo is hoping to exploit.

In the case of an earthquake, Docomo hopes it will be able to see where transit routes have been damaged and, based on historical tracking, be able to estimate the numbers of people unable to get home. For urban planning the idea is to be able to track how the distribution of people is changing over time, enabling city planners to address needs before they become problems.

The enormous amount of data stored by network operators isn't really utilised by anyone, at least not yet. That's partly the quantity of data is overwhelming (T-Mobile, for example, had 35,000 points of data for one customer), but also because of privacy concerns.

But it is interesting to note that every time there is a march, protest or gathering the mobile network operators know exactly how many people attended – even if the police and organisers can never seem to agree.

NTT Docomo won't be sharing its data willy nilly, telling Penn Olson that "We are willing to offer and make use of our operational data, if it's for our society to grow further. But it is not yet something we will open to the public in a specific format."

There's a huge amount of such data, and we're only starting to work out how best to make use of it. ®