Tomorrow's cities: Are your shoes giving away data?
Shops and retailers are taking over where street cameras left off, watching shoppers' every move.
According to a 2015 survey of 150 retail executives from IT services firm Computer Services Corporation, a quarter of all British shops and 59% of fashion retailers use facial recognition software.
Such technology is vital as offline stores attempt to keep up with online retailers, said Duncan Mann, chief operating officer at retail analysis firm Hoxton Analytics.
"Online retailers gather all kinds of information about shoppers and physical stores also want to understand how people behave in a shop," he said.
But, he admits: "A lot of these technologies are kind of invasive."
Hoxton has come up with a novel way of measuring footfall - literally by filming people's shoes.
Sherlock Holmes-like, its system can deduce a remarkable amount of information such as age, gender and social class of shoppers from their footwear.
"We have cameras at about 50cm off the ground and it points down so it is less invasive than facial recognition," he explains.
It is surprisingly accurate. It spots the correct gender 80% of the time, better than some facial recognition technologies, according to Mr Mann.
Cities are getting fuller - 70% of the world's population will live in urban areas by 2050 - but don't think for a minute that means you will be able to get lost in the crowd.
Because those cities are most likely watching you. Plugged into the network - via smart CCTV cameras that feed into central operation centres or smart street lights that turn brighter when someone walks beneath them - cities are increasingly collecting data on their inhabitants.
The purpose is to keep people safe, provide more efficient services and prevent overcrowding or other disasters but has anyone ever asked its citizens whether they want to be part of the urban efficiency experiment or offered them ways to opt out of the networked city?
"Very few of us have any real concept of what data smart cities are gathering," said Renate Samson of privacy watchdog Big Brother Watch.
"Some of it may be completely anodyne and simply a reaction to a physical movement, but with the increase of devices connected to the internet, the chances are that street lamp, CCTV camera, wi-fi connection, electronic keypad, touch and go payment device, is capturing data on you, your movements, device data and personal information."
You begin leaking data as soon as you wake up. Maybe you check Twitter before you leave the house to find out if your train is running - that tweet immediately becomes public property - or perhaps you are signed up to navigation apps such as Waze, that crowdsource real-time information about problems on the roads.
Apps collect and share lots of personal information and, in its privacy guide, Waze states that it will collect periodically "all of the phone numbers which are stored on your device's phonebook", as part of a feature to connect you to your friends.
And once you enter the transport system you are giving away even more details about yourself. The ticket gate is waiting to swallow your data, via the swipe of a smart card, mobile phone or credit card.
Transport for London (TfL) now has a bird's eye view of the estimated 4.1 million journeys taken on its network each day.
It knows where people get on and off and it can start to see patterns in the data - for instance, someone who uses the system during the day but not at peak times is likely to be a student or a retired person, someone who has one day a week when they don't use the network may work from home that day, someone who takes a brief diversion along their usual route may be dropping off a child at nursery.
"The data can be used to inform future expansion, whether we need to add a bus route or increase the frequency of trains, to alleviate capacity issues by informing people about the most crowded times and places and generally helps us to understand customers better," said Gabriel Goulet Langlois, a data scientist at TfL.
It can offer automatic refunds if people accidentally tap out with their Oyster card and could be used to alert customers when stations are closed or overcrowded.
TFL data was instrumental in solving one of the most notorious crimes of the decade - the Hatton Garden jewellery robbery - when an Oyster pass was found in one of the suspect's wallet, throwing light on how the heist was planned.
Mr Langlois is keen to point out that, while the police might want to track individual journeys, TFL does not.
"We only want to see aggregated patterns," he said.
He has been working on a project that can classify travellers, spotting commuters, visitors and even students by their travel patterns.
Big data in cities has to be handled increasingly sensitively, thinks Nick Millman, managing director at consultancy firm Accenture.
He cites the example of Google which is currently using Google Map data to assess the traffic flow in Stockholm as part of its Better Cities programme.
"It is using what is known as differential privacy," he explained. "It is basically adding privacy controls to statistics so that you only see the data you need to know about.
In the case of Stockholm, that means sufficient data to improve traffic but not so much that it reveals individual journey patterns.
CCTV cameras are a much more obvious city spy and they now dominate the urban landscape. Numbers from security firm IHS suggest that there were 245 million professionally installed video surveillance cameras active and operational globally in 2014 as more cities turn to monitoring technology to help reduce crime and anti-social behaviour.
Security firm Thales has installed 15,000 security cameras in Mexico City since 2009, which feed into operation centres around the city, including pop-up mobile surveillance centres for big events. Thales boast that it is the world's "most ambitious urban security project " and claims that since installation, it has seen
- Crime rates reduced by 48.9%
- Average response times down from 12 to 2:09 minutes
- Insurance premiums down by 30%
- 50% of stolen cars recovered within 72 hours (thanks to number plate reading technology)
But in the UK, some local authorities are scaling back on their CCTV usage, according to a recent report from Big Brother Watch.
"More and more councils are scrapping their CCTV systems due to budgetary constraints. In some cases councils have removed CCTV because the systems have no longer been necessary, crime has either moved to another area or the cameras have been of such poor quality they haven't provided any substantial benefit," said Ms Samson.