The Inquirer-Home

Police fingerprint plans scuppered by network

£2.3bn Airwave network flunks again
Fri Nov 07 2008, 14:10

PLANS TO ENDOW UK police with mobile biometric scanners that identify people on the street have hit a problem because the police network isn't good enough to accommodate them.

Airwave, the £2.3bn radio network installed for the police in 2005, was sold on the promise it could handle voice and data communications for police forces across the country for the next 20 years.

But the National Police Improvement Agency (NPIA) has admitted it is trying to acquire another network to handle data because Airwave doesn't have the capacity.

Geoff Whittacker, chief technology officer of the NPIA, told the INQUIRER, " My understanding is Airwave was not designed to take large amounts of image data. That's not what it was procured for. So its not really a suitable carrier. "

The £30-£40m project Midas, by which the NPIA hopes to distribute mobile fingerprint scanners to bobbies across the UK, has been costed to include the purchase of another data network, he said.

The NPIA invited tenders for Project Midas, the second pilot of police mobile identity scanners in England and Wales, in August. Asked by the INQUIRER whether the NPIA would have to install another secure network to support its plans, Geoff Whittacker, chief technology officer of biometrics at the NPIA, said: "This is part of the competitive dialogue. A contract will be awarded by the end of 2009."

He told the Biometrics 2008 conference last month that Project Lantern, the NPIA's first mobile biometric scanner pilot, had been forced to make do with sending fingerprint data "over a public network". The data, which amounted to fingerprints scanned by police on the beat with mobile readers, was being shuttled over "an encrypted VPN," or virtual private network, the sort of system used by private companies to handle the data communications of home workers and field sales executives.

Airwave was designed to carry police data over a private, dedicated police network to ensure the security of its data. The NPIA was keen to point out that the fingerprint data it would send over the Midas network would also be secure.

"It is a given that, as part of this information assurance process, MIDAS will use Government approved encryption products and keymat for the wireless communications, whatever the communications network," said an NPIA spokesman in an email to the INQ.

The NPIA had been forced to consider "alternatives" to the public network used by the Lantern mobile fingerprint trial, said Whittacker. "[There's] Airwave, the secure digital police network, but that's not capable of sending large numbers of images," he said.

Airwave, which runs on Motorola's Terrestrial Trunked Radio (Tetra) technology, was procured by the Police Information Technology Organisation (PITO), the NPIA's precursor, in 2000. It wasn't fully operational until 2005 and was projected to cost £2.3bn over 22 years at 1999 prices. London's Metropolitan police hadn't even finished introducing Airwave devices to all its 27 boroughs last year.

Macquarie, an investment fund, bought Airwave for £1.9 billion last April with the declared intention to run it for at least 10 years, some five years short of its original lifecycle. The system is used by 180,000 police officers, including those in the Serious and Organised Crime Agency, for whom security of communications is paramount.

The NPIA had sold Airwave to national police forces on the promise that it could "transmit voice and data communications at the same time". It was also celebrated for its "Improved security and encryption of communications".

Individual police forces had to buy their own Airwave radios, and modernise their control rooms to handle the digital communications. The NPIA did not say in its promotional materials what the this additional cost was to the UK tax payer, merely that it was funded by the Home Office and through local Police Authority funds raised through the Council Tax.

Whittacker said suppliers tendering to run Midas would be asked to come up with ideas to solve the police network problem, so the options were still open. The standard method for capturing fingerprints with mobile readers was to scan an image and then send it over a network to a computer where it was processed. Airwave couldn't handle this, but it might still be possible to find a way to use the £2.3bn network. It might be possible, for example, to invest more money in scanners that can process the fingerprint images and turn them into shorthand codes that can be sent over the network instead.

"All of these options are open at the moment and to start discussing any particular one may be tricky, including Airwave, so we don't really want to start going into specifics," said Whittacker.

However, the NPIA also wanted Midas to handle other biometric data, said Whittacker: "We don't want to constrain ourselves to fingerprints". Other options for the mobile police scanner would be to capture mugshots and DNA. The police would not be able to identify people with their mugshots yet because there is not a national mughsot database. The overarching brief was "mobile ID" , he said.

Another sales job
After convincing police forces to invest in Airwave, the NPIA must now embark an another sales job, to convince national police forces to buy its biometric scanners and another network to support them. They would also have to invest in further changes to their control rooms to accommodate the Midas network.

VPN technology like that used by the Lantern mobile pilot would be significantly cheaper than Airwave, but the NPIA has not published the results of the Lantern pilot, so it is not possible to tell how secure it was. There is no assessesment report for Lantern. Nor has the NPIA published a business case for the £40m Midas procurement it opened in the summer.

Regarding the network Lantern used as an alternative to Airwave, Whittacker only said it had not been good enough. But, said the NPIA press office, "part of the purpose of the Lantern pilot" had been "exploring and testing communication networks".

The NPIA is using statistics taken from the Lantern pilot to justify its expenditure on Midas. Hence, from the perspective of the NPIA, the quango that has both its reputation and future income riding on the success of its biometric technology venture, Lantern, and therefore the case for Midas, is shining like a brass knob.

Whittacker's presentation on Midas last month was full of promise for the new technology. Project Lantern, which was intended as a "proof of concept" pilot for police fingerprint readers, had been "a huge success", he said, adding, " It's fulfilled its purpose and demonstrated the benefits quite clearly."

The NPIA has also not stated how much of the money invested in Airwave would now be wasted because its facility to handle data was already redundant.

But Whittacker did say that Lantern, which distributed 200 mobile fingerprint readers to 20 police forces, had proved that fingerprints could be retrieved from the police force's IDENT-1 fingerprint database within two minutes, in 80 per cent of cases.

The pilot readers had been distributed to police units that were also testing Auto-Number Plate Recognition cameras. It worked on the assumption that people whose car licence plates drew warnings from the Driver and Vehicle Licensing Agency (DVLA) database were probably also likely to be involved in serious crime. The automated checks drew warnings for things ranging from speeding offences through driving without insurance to possessing a stolen vehicle.

Thus when Whittacker told the conference last week that "60 per cent of drivers stopped as a result of ANPR, give or attempt to give false details to the police", he was presenting a justification for the fingerprinting exercise. But an objective assessment of the Lantern pilot would give an indication of the number of unsolved serious crimes that have been committed by those drivers flagged up by mobile police ANPR cameras and subsequently indentified by mobile police fingerprint scanners.

Should the Lantern pilot be proven mostly as a high-tech means of catching people without car insurance, the £40m Midas might not be easily justified. The police were seizing 1,500 uninsured vehicles a week, Rosie Winterton, the then minister for state at the Department of Transport told a parliamentary debate last July. These seizures were made without the help of fingerprint scanners.

Indeed, the government has already invested a considerable amount of money in the means to catch problem drivers like those without insurance, including the introduction of ANPR intercept teams in every basic command unit of the police (of which there are several in each of the 43 forces in England and Wales), communications links that allow ANPR cameras to access databases with vehicle information, and a national ANPR data centre capable of storing 50 million number plates a day and storing them for five years.

Taking liberties
A more compelling reason for police forces to buy the NPIA's Midas sales pitch is its ability to improve police intelligence on the beat. It helps prevent wanted suspects from evading the police by giving them false identities. Normally, if a police officer has sound reason to suspect someone, they must take them to a police station to have them fingerprinted properly with ink rolls. Images of those rolls can then be used to make relatively reliable matches against the police database.

By allowing police to take cursory electronic fingerprint scans while out on the beat, Midas would reduce the numbers of people who are pulled in unecessarily for a lengthy fingerprint session at the station. This, said Whittacker, result in more public confidence, less bureacracy, fewer unlawful arrests and a reduction in the number of ID arrests, which are the occassions when the police take people to the station just so they can be ID'd.

"The information we've managed to get so far is we've saved 67 minutes per [fingerprint] search," said Whittacker. "That equates to 366 more police officers on the beat if scaled up," he said.

The scanners could be useful for "non-designated rural police stations with no livescan fingerprint capability", he said. They could also be used by police at public events, during special operations and to stop prisoners swapping their identities in transit.

Whittacker did not report what the trial said about the quality of the responses to the mobile fingerpint searches. The Lantern searches, of which 30,000 were made, alerted police officers when a quality threshold had been reached on a match against the UK's IDENT-1 police fingerprint database.

Possible matches were reported with "high or medium confidence" and up to three identities were suggested as matches against the IDENT-1 database. Police officers would take people to the fingerprint suite of a major police station if they thought they needed to do a full fingerprint seach with rolled (inked) prints.

The accuracy of the system might be important, particularly as it involves people's police records. In last July's Parliamentary debate on the Motor Insurance Database (MID), which is referenced by police ANPR scanners, David Jones, Conservative MP for West Clwyd, said that as there were 33 million insured vehicles on British roads, and the MID had 5.6 per cent level of inaccuracy, 1.5 million insurance records used by the police might be inaccurate.

"The consequences of such inaccuracy can be serious," he said, before giving some examples of just how serious it could be.

Maureen Smith, of Rhos-on-Sea had lent her car to her 29-year-old daughter, Helen Parry, who had her own, portable insurance. The database told North Wales police that Parry wasn't insured to drive the car. Parry explained that the police seized the car and left her standing in the rain, waiting for a lift from her mum. She was 20 weeks pregnant. Mum came to the rescue, but the police didn't believe her either. The police took Mum to the station and said they wouldn't return her car until she paid a £105 fine. The police made up for it later by cancelling the fine and buying flowers. Mum's insurer's, Direct Line, had failed to update the MID with the details of 400,000 insured drivers.

Another example given by David Jones MP was that of Jane Groves, who changed her insurer and put her partner on her policy. The police stopped her partner. The database said merely that the old insurance policy had been cancelled. Grove's car was seized and she was charged £105, which had not been returned at the time Jones recounted the story.

"The database is an essential component in the drive to rid our roads of uninsured drivers. They are a menace to others, and they cost honest motorists hundreds of millions of pounds per annum," said Jones.

But he added: "The problem lies not so much with the database, as with the way in which it is used by the police, and the promptness with which insurers update it. It is accepted that the database will always be less than 100 per cent accurate; that must be the case, and there are many reasons why."

The police had been heavy-handed, he said, and driven by a target culture that almost encouraged them to sieze vehicles and issue fines.

"The database, in short, should be treated as a very useful tool, but not as infallible. Police officers should be told of the need to use it with discretion and common sense. If police officers do not use discretion, injustices of the sort experienced by Mrs Smith, Mrs Parry, Mrs Groves and, I have no doubt, many others will continue to occur," said Jones.

So error-prone databases are used to feed error-prone ANPR cameras, which are used to prompt error-prone fingerprint checks.

The NPIA's sales pitch for Midas didn't include anything about this. µ


Share this:

blog comments powered by Disqus
Subscribe to INQ newsletters

Sign up for INQbot – a weekly roundup of the best from the INQ

INQ Poll

Heartbleed bug discovered in OpenSSL

Have you reacted to Heartbleed?