Along with the publication of the new version of the controversial Identity Cards Bill yesterday, the Home Office also revealed some details of the UK Passport Office Biometric Trial.
The Home Office now seem to have censored their Press Release, which now no longer states right at the top:
"Please Note: There are currently no plans for further biometric trials."
Surely they should have issued a separate press release or statement about this, rather than just altering the original online version ?
The Press Release points vaguely to the UK Passport Service website for the Report and Management Summary documents about the Biometric Trial, which can now be found through links on this page.
Upon reading the reports, it swiftly becomes clear why the results of this UK Passport Office/Service Biometric Trial project, which after months of delay and massive initial technical problems finished in December 2004, seem to have been supressed during the debates on the Identity Cards Bill earlier this year, and during the General Election - Biometrics simply does not work swiftly or accurately enough for a massive National ID scheme involving a population of 60 million people.
How can an average Verification delay of between 40 seconds, for the best group and 80 seconds, for the worst performing group, even under ideal lighting conditions, be anywhere near to acceptable performance for a system which is intended for queues of hundreds of passengers at airport Passport Control ?
There were cases of Verification taking on average 10 minutes per person.
Anything that adds this amount of extra delay to the existing Passport Control procedures of forming a queue, fumbling with documents, having them examined etc. will bankrupt British Airports Authority and the airline and travel industries.
If Biometric checks are mostly not performed on British citizens and residents as they enter the UK, then what is the point of claiming that Biometric Passports are needed in the first place ?
The reported error rates are also astonishingly bad, ranging from 4% for iris scans to around 30% failure for facial recognition.
For a workable biometric system suitable for use by millions of people, the error rates have got to be orders of magnitude better than this i.e. fractions of a percent, otherwise there is no hope of ever catching any significant number of wanted terrorists or illegal immigrants, which is what the Government claims their complicated and costly ID cards/ Centralised Biometric Database / "e-borders" plans are meant to do.
Neither of these biometrics are mandated for by the ICAO machine readable travel document standard, which simply states that a digitised photo stored in the passport, which can be manually compared on screen with your face by a human passport control official, with the option of an as yet undetermined number of fingerprints,
This Passport Office Biometric trial was originally inteded to test Biometrics with a view to creating a passport ID card acceptable for travel within the European Union. The terms of reference of this trial were only to test business processes and procedures during enrolment and registration onto such a system. It specifically was not budgeted to test any of the large scale infrastructure preformance problems or to conduct any kind of security penetration testing to see how vulnerable the system is to various Biometric Identifier snarfing, interception, cloning , replay, network infrastructure compromise, database compromise and Denial of Service etc attacks which both a Passport and an National ID system have to be able to withstand with ease.
It is true, therefore that the results from the particular hardware and software used in this trial are not necessarily a reflection of the performance that may be achievable in the future.
However, there is no question that this snapshot of the current state of the biometric techhnology art is simply not up to the job being asked of it.
Unless the Home Office repeats this Biometric Trial, every few months, into the future , until acceptable levels of accuracy and speed of throughput are regularly achievable with commercial off the shelf equipment, then Yet Another Government IT Disaster looms
for Biometric Passports, let alone for the wider use of such technology, under more adverse and varied conditions, for a National ID card scheme.
However, it would appear that the NuLabour governnment plans to press on regardless, and that the politicians and spin doctors still cling to the absurd notion that somehow biometric technology is an unforgeable magic fix to the ever changing list of social problems (this week it seems to "Identity theft", during the General Election campaign less than a month ago it was "terrorism") that the ID cards scheme is alleged to be a cost effective soltion to.
Leave a comment