WRITTEN ON May 26th, 2005 BY William Heath AND STORED IN Identity
ITV news reported the Passport Service biometric trials had 4% failure rate for iris, rather more for fingerprint, and a whopping 31% failure rate for facial recognition. Perhaps they should have added tossing a coin. It’d be cheaper, and right 50% of the time.
This important report by Atos Origin is not easy to find, but the crafty Spy blog* points out there’s a summary at the foot of UKPA’s publications page, or you can now download it here. *Cheers for that.
5 Responses to “Heads I’m who I say I am, tails I’m not”
The report and management summary are hiding under the Publications section of the UK Passport Service website:
The Verification times, under the ideal conditions of the enrolment centres, are astonishingly slow.
40 to 80 seconds average delay per person, when you are at the back of a queue of 400 Jumbo jet passengers is going to lead to intolerable queues, and the financial ruin of British Airports Authority etc.
The worst case verifications took, on average 10 minutes each !
Any real life deployment of such “state of the art” equipment under less than ideal lighting conditions e.g. a local bank or doctors surgery is going to be even worse.
The error rates are also way too high – they need to be fractions of a percent, not 4% to 30% !
The Government will spin that this trial was not actually testing the Biometric technology per se, but then it was not testing any of the large scale infrastructure or security threats either.
Astonishingly, the Home Office Press release
http://www.homeoffice.gov.uk/n_story.asp?item_id=1303
states right at the top:
“Please Note: There are currently no plans for further biometric trials.”
Of course Atos Origin, like any good consultamnts obviously reccommended “further studies are required”.
The Home Office has now shoved the report up here:
http://uk.sitestat.com/homeoffice/homeoffice/s?docs4.UKPS_Biometrics_Enrolment_Trial_Report&ns_type=pdf&ns_url=http://www.homeoffice.gov.uk/docs4/UKPS_Biometrics_Enrolment_Trial_Report.pdf
The Home Office now seem to have censored their Press Release:
http://www.homeoffice.gov.uk/n_story.asp?item_id=1303
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 ?
Weird about the iris failure rate. I went to a lecture by John Daughman http://www.cl.cam.ac.uk/users/jgd1000/iris_recognition.html which said no false positives have occurred in the UAE which use iris recognition for all immigration. “This is the reason why the United Arab Emirates deployment can perform 3 Billion iris comparisons every day without getting False Matches.” see article http://www.cl.cam.ac.uk/users/jgd1000/largedatabases.html