FEATURE Rising awareness of the benefits of AFIS technology has had a positive impact on the industry. Wider deployment of credible reference sites has further attributed to the market positioning of AFIS technology. However, several industry challenges need to be addressed. Government-backed projects typically have long decision cycles that can sometimes take over a year, depending on the size and complexity of the project. Poor economic conditions in some countries have resulted in lack of availability of financial resources. Privacy concerns pose barriers to civil application projects. Proposals to issue national identity cards in some countries are met with severe opposition from various civil liberties and privacy concern groups.
and their applications much more economical to install. Large-scale companies should look for opportunities for mergers and acquisition in order to establish a strong foothold in the industry and provide the end users with a range of products and services. For small and medium sized companies this is the right time for consolidation of resources and target new areas of opportunities.
Standardisation
“Though prices have declined considerably because of the competition, cost has remained a challenge. The ability of participants to develop and introduce new capabilities at competitive prices is critical”
Standardisation has become critical for user adoption of AFIS. Delay in adhering to these standards by vendors has hindered the growth of the AFIS market. For standards to be effective across borders there is a need for participation from vendors worldwide. Though prices have declined considerably because of the competition, cost has remained a challenge. The ability of participants to develop and introduce new capabilities at competitive prices is critical. Falling prices of security products are likely to increase adoption during the next two to three years and will make the security systems
In the past three years, there have been some major acquisitions, and it is expected that biometrics industry will continue witnessing consolidation in the short and medium terms with smaller companies being acquired or merging with larger companies to gain economies of scale. Additionally, it will benefit end users by providing them with a cost-effective, broad range of product portfolio. In July 2011, Safran finalised the acquisition of L-1 Identity Solutions. This acquisition has strengthened Safran’s position in the biometrics market. Safran had previously acquired
Motorola’s biometric business unit, including the Printrak trademark. In August 2010, 3M entered into an agreement to acquire Cogent in order to expand its reach into access control and other commercial ID and authentication applications. Within the realm of security, technology is being recognised as a tool to assist in the detection of terrorists and prevention of terrorism. Eager to capitalise on the tremendous opportunities opened for security products, many biometric companies are moving to capitalise on security by partnering, acquiring, and merging with security companies. Hence, small-scale companies should form strategic alliances and large-scale companies should look for acquisitions in order to progress in a challenging market. National ID and e-passport projects, seaports, airports, and other infrastructure security will continue to drive revenues for vendors in 2012. Their focus on innovative solutions, higher performance levels, broad product portfolio, and value added customer service will be key for winning contracts.
About the author Neelima Sagar is senior research analyst, Automatic Identification and Data Capture, at Frost & Sullivan North America and India. Her experience covers a broad range of automatic identification technologies across all global sectors. Areas of expertise include the biometrics and RFID industries.
Document quality issues jeopardise border biometrics Georg Hasse, Secunet Security Networks AG With an ever increasing number of travellers taking to the skies, governments are working to ensure their border control procedures meet the growing demand for better security and speedier throughput. According to international aviation trade body IATA, airline industry revenues from passenger travel grew from US$23bn in 2001 to an estimated US$456bn in 2011. While the figures make good reading for the travel and tourism industries, they create real challenges for border control authorities around the world, who must make travelling as pleasant, swift and secure as possible for all travellers. Secure and efficient border control is an economic necessity because our airports play
January 2012
an important role in keeping the wheels of the global economy turning. For example, if a major airport gets a reputation for slow border throughput and inadequate facilities, that could affect a company’s decisions to invest in that location if it relies on international travel for business. Equally, travellers will vote with their feet and opt for a rival location if they experience any inconvenience at an airport. As a result, airport authorities are investing heavily in automated and
semi-automated border travel solutions based on face recognition technology, that speed up passenger throughput without having an impact on the effectiveness of security procedures.
Considering process There is no doubt that the latest border control systems have the potential to improve processes for border officials. However, it is not enough simply to implement a technical solution and then expect it to work. In fact, every part of the process has to be considered in detail, from how travellers enrol in the system to how it will actually operate.
Biometric Technology Today
7
FEATURE
The quality issue
European e-borders: data quality on e-passports is an issue.
Travellers should be able to follow instructions at an automated border control (ABC) and present their ID document – such as e-ID card or e-Passport – to the machine for checking. Unfortunately, some countries are now finding to their cost that the system doesn’t work as well as expected. Certain nations have chosen to implement e-Passports and have not ensured good data quality. Most countries introduce e-Passports for a good reason – to enable machine-assisted ID verification. But problems are introduced into the system during the enrolment process because not enough care is taken with the quality of the data. This has led some industry insiders to question the point of implementing e-Passports, when a poor implementation adds significantly to the cost.
“Most countries introduce e-Passports for a good reason – to enable machineassisted ID verification. But problems are introduced into the system during the enrolment process because not enough care is taken with the quality of the data” As with many data solutions, it’s not the technology that’s at fault, but the quality of data. The IT world is littered with experiences of Garbage In Garbage Out (GIGO) and the border control world is now falling victim. Essentially, not all of the new e-Passport and e-ID documents are up to the job. Although the technology itself works properly, existing quality control issues in some country’s documents have the potential to cause real headaches for border officials in the future. For example, Frankfurt’s eGate EasyPASS ABC system that works with EU e-Passports has recorded a sizeable number of people being turned back from the border control gates because the quality of the biometric data in their ID document is not high enough. 8
Biometric Technology Today
In spite of the raft of e-Passport standards and specifications such as ICAO doc 9303, some governments are failing to issue robust documents that meet the needs of border control systems. And, because a government-issued ID document is used not just in the document holder’s country of origin but also abroad, it’s even more important to ensure high quality biometric data and enrolment that meet recognised international standards. For example, photos used in e-Passports should comply with ISO/IEC 19794-5, which describes interchange formats for several types of biometric data. It also defines a standard scheme for codifying data describing human faces within a CBEFF-compliant data structure, for use in facial recognition systems. The standard is intended to allow computer analysis of face images for automated 1:n matching and 1:1 authentication, as well as human identification of distinctive features such as moles and scars. To enable applications to run on a range of devices and to improve accuracy, the specification also describes an array of other requirements such as lighting, pose and expression, positioning, image resolution and image size. In addition, it details what should be included in the Full Frontal Image Type applied to e-ID and e-Passport documents, which in this case specifies the capture of the full head of each individual.
German process Although Germany is renowned for its highly secure ID documents, even it has had cases where the application process has impaired the system. For example, to obtain an e-ID, citizens have to visit their local council and present a headshot, which may or may not have been taken by a professional who’s familiar with the technical guidelines. The council then glues this image to the application form, scans it in and sends it off to the secure printers. The problem with this approach is that image quality control is invariably inferior and the individual registering for the document may find it does not work when they are in a hurry and trying to pass through border control at an airport. So, although Germany does have quality control guidelines for facial images that are being used for e-ID documents, it has problems with its processes. The citizen only finds out whether or not the facial image they brought with them is suitable when they hand it over to the council official.
Live enrolment techniques, in contrast, would provide instant feedback and there would be no need for citizens to bring along a printed image. A pilot project in Monheim, a city close to Cologne, is using a live enrolment solution.
“Although Germany does have quality control guidelines for facial images that are being used for eID documents, it has problems with its processes” Germany’s application and image-capture system is not alone in its shortcomings. Another western European country is known to issue a passport from every application it receives, regardless of the photo quality. This is because in the agreement between the document supplier and the national authority, the responsibility for quality control lies with the national authority, and the printer carries out no additional checks. Although this saves money, it may not be cost effective andraises questions about whether local authorities care about data quality. Additionally, the country in question doesn’t want to compete with local photographers, so even though it is now sourcing enrolment equipment, it will keep accepting printed photos from applicants. This is always a security issue since it’s hard to tell if the picture has been modified or even shows a different (but similar) person. Any passport manufacturer that has a 0% photo rejection rate should examine its quality control procedures because it is highly unlikely that every citizen will provide a perfect image that conforms with the ICAO’s recommendations. However, it should also be remembered that even with good quality control during enrolment, photo quality might be compromised through generation losses in print/scan cycles. When using paper photos submitted by the citizen, rejection rates of up to 10% are not uncommon. With live enrolment and ICAO quality control, the rejection rate is brought to fractions of a percent, while saving citizens time and governmental money. Local administrative bodies that collect biometric data should also be given clear guidelines and education about how to check image quality, as well as ensuring that the photo standards they issue are ICAO-compliant. This will ensure citizens get the valuable document they paid for and prevent them from demanding their money back when a poorly produced e-Passport is rejected by a border control system.
January 2012
FEATURE
Value for money All governments want to minimise costs when it comes to spending taxpayers’ money. But it is important to remember that value for money is just as important as keeping costs down. Sometimes it makes sense to spend more money upfront to ensure less long-term expenditure, and this is certainly the case for e-Passports and e-IDs. It is a false economy not to invest in quality control at the outset, because documents that are not fit for purpose will end up being rejected before eventually having to be re-issued at a cost to both the issuer and the document holder.
"It is a false economy not to invest in quality control at the outset, because documents that are not fit for purpose will end up being rejected before eventually having to be re-issued at a cost to both the issuer and the document holder" If authorities have invested huge amounts of money in passport production systems and fingerprint enrolment stations, but haven’t thought about what the digital picture in the document will be used for or whether it will even work, then they could face significant future costs to put it right. In other words, it is essential to analyse the total cost of ownership. When issuing a tender for passport suppliers it is vital to understand exactly how the passport will be used in the future. If it is likely to be used for computer matching at automatic border controls – and there is a good chance that this will happen as it is a growing trend – you need to be sure that the document will work. The German BSI has issued technical guidelines covering this aspect of issuance. Luxembourg tendered centrally and then, in 2009, rolled out a uniform Speed-Identity solution to all municipalities responsible for enrolling citizens for passports but not sourcing the equipment. The country is using live enrolment including instant quality assurance to ensure ICAO-compliant face and fingerprint capture, and has achieved very good results. Furthermore, governments should ensure that they include in the decision-making process all organisations that will use the document. If the same enrolment equipment can be used for e-Passports, e-IDs or e-Visas, it is possible to achieve economies of scale.
January 2012
Governments keen to get value for money may consider a public/private partnership, or a payper-document or pay-per-use model. These have the advantage of ensuring that the governments themselves have a high quality enrolment solution but don’t have to bear all the costs at the outset. This ‘identity as a service’ concept means that instead of investing in equipment that it would then own, they can buy a service from local providers or systems integrators. This approach is already gaining traction, as the private supplier would have the same incentive as the government to keep the equipment up and running, because otherwise it wouldn’t get paid. However, it should be remembered that such a service requires a local service set-up with very short response times. National administrations that have invested in well defined, high quality enrolment solutions should also bear in mind that there will always be a proportion of the population that can’t make it to a local council office, embassy or consulate to register.
to get their fingerprints captured or have their photograph taken again. The lifespan of e-Documents should also be examined. Some countries are considering introducing biometric passports, which are valid for only five years rather than the usual 10 because people’s faces change and more regular image updates improve the computer matching process. Interestingly, Sweden went from 10 to five years in 2005 after it had introduced e-Passports. Others remain reluctant to take this route because they believe that reducing the lifespan will be too costly. However, countries that opt for a document lifespan of 10 years should bear in mind that a five-year validity is better for quality control purposes; this has been recognised by the ICAO, which recommends a maximum document validity of five years. In fact, because of the quality control issues, it’s debatable whether anyone will benefit or any savings will be made from sticking to a 10-year lifespan.
“Countries that opt for a document lifespan of 10 years should bear in mind that a five-year validity is better for quality control purposes; this has been recognised by the ICAO, which recommends a maximum document validity of five years”
The 2011 crisis in the Middle East and North Africa has sparked an increase in migrants from the region, and governments throughout the EU have been examining how they can respond to the pressure this is putting on their border controls. The open border policy of the Schengen region is being questioned, which is putting pressure on governments that signed up to the original free-movement agreement. When you have had limited border policies with your Schengen partners for more than 15 years, it is difficult to increase and manage new border controls overnight. E-gate solutions seem the likely way to go as they are cheaper to implement and also require a smaller workforce to be recruited and trained in border control procedures. Yet, while e-gates are the future, they will only be truly successful once governments improve their document personalisation quality control.
Consequently, they may consider deploying a mobile solution that takes document registration to more convenient locations. Countries with remote populations or whose citizens are spread throughout many locations worldwide could offer this type of system, which travels nationwide or even internationally to give everyone the opportunity to enrol. For example, Swedish embassy consular staff in the US are using mobile solutions to improve services for Swedish citizens living in the US by regularly travelling with a mobile enrolment kit to consulates and temporary enrolment offices, often in cooperation with the Swedish church. How to integrate quality controls at the time of enrolment is another area that needs to be considered. This provides assurance that the enrolled data is already quality controlled when documents are sent off for personalisation, thereby obviating the need to call citizens back to the local council offices or police stations
A question of control
Resources •
•
German BSI technical guideline for tenders: https://www.bsi.bund.de/ContentBSI/ EN/Publications/Techguidelines/techniguidelines.html ICAO standards: http://bit.ly/yUIm4k
About the author Georg Hasse is a senior consultant at Secunet Security Networks AG. Previously he held a number of senior positions at Cross Match Technologies.
Biometric Technology Today
9