Examination of Witnesses (Questions 300-319)
MS KATHERINE
COURTNEY, DR
HENRY BLOOMFIELD,
MR NIGEL
SEED AND
MR MAREK
REJMAN-GREENE
22 MARCH 2006
Q300 Adam Afriyie: At the moment
it is speculative, do you need to do further work before these
numbers and statistics are clarified?
Ms Courtney: We have always said
that we would do that testing as part of the procurement.
Q301 Adam Afriyie: We have had chip
and pin, multi-modal biometrics (face/iris/fingerprints), testing
systems, enrolment systems and verification checks. These are
all to do with technology. What are the known limitations in the
proposed scheme and how are you looking to address them?
Ms Courtney: Known limitations
in respect of?
Q302 Adam Afriyie: In respect of
the technologies that you are proposing at the moment or the route
that you are taking—the plan that you have; there must be
some known limitations with the technology and known limitations
with the schemes with the tests that have been undertaken. What
are they?
Ms Courtney: In terms of our delivery
risks, which includes obviously any technical implementation risks
but also it is quite importantly focused on how we organise the
services, how we design the business processes, how we operate
them in practice, I think Dr Bloomfield has spoken about the universality
of some biometrics. We know that there are limitations. You cannot
record someone's fingerprints if they do not have any fingers.
That is a known limitation and one of the reasons behind our intention
to use multiple biometrics to try to overcome that limitation.
The biggest risk obviously in any business process is that you
do not train your people appropriately. Because we are implementing
this with the intention of creating an organisation based on the
Passport Service, building from the good operational track record
of the Passport Service in recent years, we have every confidence
that we will be able to have the right training in place for people
so that we can overcome that possible limitation.
Q303 Adam Afriyie: In your evidence
you acknowledge that the field is fast-moving. Have you made any
projections about how technology will change over the next several
years during the testing and deployment of the project? If so,
what are the changes that you envisage? Have you planned to incorporate
them into the scheme that you are putting forward at the moment?
Ms Courtney: I might ask Marek
to speak a bit about how the biometrics field is moving forward
and also Nigel afterwards to say a few words about how we are
building that sort of flexibility into our requirements.
Mr Rejman-Greene: You made a good
point there inasmuch as the results certainly of the feasibility
study in 2002. The experience in the United States with the US
visa programme was based on technology which is now quite a few
years old. We know, for example in the United Arab Emirates, that
there is now a programme using IRIS for nearly one million people.
There is beginning to be not only an advance in the technology—the
matching and the actual sensors that are picking up the fingerprints
and the iris patterns in better and more inclusive ways and more
and more people are being enrolled and some limitations are being
countered—but the experience in terms of the larger programmes
abroad is also bringing in knowledge. The future developments
that we are foreseeing, certainly in terms of multimodal fusion
which you mentioned, means that there is a lot of research work
going on there. During the course of the deployment and early
years of the programme, we would certainly ensure and ask the
consortium that was winning the project to take advantage of that
knowledge and home in on it.
Q304 Adam Afriyie: It certainly sounds
as though the project you are proposing means that we are going
to be the pioneers; we are going to be at the leading or cutting
edge rather than adopting systems which are fully tried and tested
in the way they are going to be used.
Mr Rejman-Greene: We are co-ordinating
all those technologies, yes, but individually all those technologies
are being used and being developed in single trials. I think the
idea about actually working through multiple technologies is perhaps
the novel element in this area in order, as Katherine said, to
ensure that the highest proportion of people are "enrollable"
in the system.
Q305 Dr Turner: What about the security
of the system? What steps will you be taking to guard against
falsification of biometrics, and perhaps the most extreme case
one could imagine is that al-Quaeda would become very sophisticated
and hack into your database and plant completely false biometrics
for a different individual. What steps are you taking to ensure
the security of the system?
Ms Courtney: From the beginning
when I joined this programme, I was intent on having the best
security advice possible, and so we brought in not only the government
security advisers but also other independent security advisers
to work with us on this. Before we had a reference solution, when
we were just thinking about the principles of the scheme and the
policy decisions around that, we had security advisers alongside
us looking at all the possible risks of the scheme. We have had
that built into our design from the beginning. We asked a long
time ago for this whole scheme to be certified as part of the
critical national infrastructure. It does not exist yet, but already
it is listed as part of our critical national infrastructure and
so it is being accredited by the government's security advisers,
security accreditors, from its earliest inception.
Q306 Dr Turner: What does accreditors
actually mean?
Ms Courtney: If you would like
a practical example of that—
Q307 Dr Turner: Is it a kind of kite
mark?
Ms Courtney: It is a bit more
in depth than that. Perhaps I can ask Nigel to talk about the
security accreditation.
Q308 Chairman: Can I ask you not
to because we will come back to database security later. I know
my colleagues are keen to do that. Before we move off this section,
can I summarise where we are here? The Government has clear aims
in terms of what it wants biometrics to do in this programme.
You do not know, however, what the technology is because some
of it may not even be there yet; it might evolve over the coming
months, and so your specification in terms of level one procurement
is crucial in terms of setting up parameters for what the technology,
when it exists, will deliver. Am I right? Is that fair?
Ms Courtney: Yes.
Q309 Chairman: Could you either now
or in writing tell us the accuracy levels that you want for each
part of the biometrics? I understand not many people do not have
hands. We are talking about accuracy levels for, say, somebody
who is a builder and has a cut on his finger, or something of
that nature. Do you have those figures now?
Ms Courtney: I would like to offer
to write to you on that subject.
Q310 Chairman: If, during the process,
you find that blips come into the system, are you prepared to
say, "We are going to have to stop this and elongate the
time in which we can deliver"? Is that part and parcel of
your thinking?
Ms Courtney: Our plans have always
been to take an incremental implementation to this in a step-by-step
way, building on other developments and rolling out over a period
of time, I think from the very first policy announcement when
the Home Secretary was quite clear that there would be no big
bang implementation of this scheme. That gives us lots of opportunity
to test and ensure that we are getting things right. We are also
taking the whole programme obviously through the Office of Government
Commerce gateway process for every key component of the programme.
We are also running our own internal health checks. We will not
proceed to the next phase of any aspect of the programme without
a clear health check that tells us that we are ready to proceed
to the next stage.
Q311 Chairman: Is there evidence
to show that this is the best way of developing this scheme? Is
that evidence you have from elsewhere where systems have been
rushed?
Ms Courtney: There is lots of
evidence from the National Audit Office and the Office of Government
Commerce and elsewhere. We have certainly learnt lessons from
other programmes around the world.
Q312 Bob Spink: Could you tell us
how many people eventually will be using this scheme or enrolled
on the scheme in total, how many millions, and how many points
of access to the scheme checking people there will be eventually—how
many tens of thousands of those?
Ms Courtney: The expectation is
that in terms of customers or individuals enrolled in the scheme,
eventually that will reach about 60 million. We will, however,
have to hold records in the scheme on people who have left the
scheme.
Q313 Bob Spink: How many points of
access to the scheme will here be?
Ms Courtney: I would like to clarify
that we are not talking about access to the system. I think that
word is often used and misconstrued. We are talking about designing
a system here which allows people to present identity information
and have it confirmed.
Q314 Bob Spink: We are finishing
very shortly and we only have 30 minutes. Could you just address
the question specifically and, if you need to add to it, perhaps
you could write to us.
Ms Courtney: I cannot give you
a number or the volume of verification transactions that we would
expect to see on the system.
Q315 Bob Spink: At every airport
and port of entry in the country, at police stations and at social
benefit offices and so on, how many points will there be in the
country in the whole system?
Ms Courtney: We do have assumptions
around this. It would be better if I offered to write to the committee.
Q316 Bob Spink: Can you just give
us a rule of thumb now?
Ms Courtney: I will not be able
to do that.
Chairman: We are happy about your writing
back to us on that.
Q317 Bob Spink: Clearly, you said
that the success in matching was fairly high, in answer to Dr
Turner. Could you also say for each of the systems—iris,
fingerprint, face recognition—what "fairly high"
actually means? Could you write back to us on that, too?
Ms Courtney: Yes, certainly.
Q318 Bob Spink: Things are changing.
I have learnt something this morning. It is now not a fused biometric
system; it is a pick any one from three system. Do you think that
if that is the case, if it is just pick one from three and try
to match it, this will limit the ID card system's ability to deal
with immigration, crime, terrorism and ID fraud?
Dr Bloomfield: I would go back
to the conclusions of the 2003 NPL feasibility study, which recommended
that, in order to differentiate between all the individuals in
a population of 50 million, enrolling four fingerprints would
be sufficient or enrolling both irises would be sufficient. The
conclusions of that report were that you could use either four
fingerprints or two irises in order to discriminate amongst individuals
in a 50 million population. I think the answer to your question
would be no, that how we choose to combine these biometrics would
be sufficient to identify individuals from the population.
Q319 Bob Spink: In answering my colleague
who asked for the limitations, you did not say what the limitation
on iris recognition would be, for instance for women who were
in menstruation where the rejection rate increases very dramatically,
as I am sure you are understand, or on fingerprint recognition
for people who are over 60, or bank clerks or teachers where fingerprints
fail, as we saw with our Chairman who got two out of three failures
since he was a teacher on a straight one-to-one fingerprint recognition
in America a couple of weeks ago. Perhaps you will write to us
about that as well. The whole policy of ID cards is predicated
on an assumption that the technologies will work eventually: is
that true?
Ms Courtney: The decision on the
policy on ID cards was taken by the Government on the basis of
quite a lot of analysis and the technology was only one aspect
of that. Certainly the advice that we have received all along
has been that the technology will be fit for the purpose to support
the business objectives.
| 
