The most important results of the BioTrusT project

Demonstration platform for 12 biometric systems

Stargate at FH Giessen-Friedberg

  • More than 400 users, used over 75.000 times
  • Model for positioning sensors and user guidance to be used also by handicapped people

 

Test of 8 biometric systems at 4 locations*

for access control in open system environments with volunteers (ca 200 users)

  • Recommendations for biometric systems implementation and usage by a state’s privacy officer and a consumer advocate
  • General recommendations
    • User guidance strongly recommended
    • Quality of first enrollment is critical to achieve later high recognition rates; quality of enrollment systems and procedures are key
  • Other significant findings
    • With any single system some persons could not be enrolled with sufficient quality
    • During the tests enrolled persons were sporadically not recognized by almost every system (caused almost equally by either, environmental or behavioral modifications)
    • Alternate equivalent biometric systems are needed for larger groups of people – not at least due to individual preferences of specific systems
    • Standard interfaces are required to allow for different biometric systems and to protect the investment of an operator
    • Findings and usage recommendations have been implemented by project members (system providers)

* Münster, Friedberg (2), Stuttgart

 

Development and test of a standard interface for biometric systems

  • Adaptation and first of a kind implementation of the international industry standard BioAPI 1.0/1.1 in applications (e.g. screen saver[1], log-in[2], browser-plug-in[2], HBCI-home banking[3]) and biometric systems[4] on Windows 95, 98, 2K and NT
  • Decentralized test at up to 12 sites distributed over Germany with templates placed at
    • PC/Server
    • Chip Card (ITSEC E4 high, template controlled by its user)
  • Development and test of a common feedback capture and analysis system by close cooperation of research, privacy officer and consumer advocate
  • Findings
    • Complexity of BioAPI standard is high, integration with chip card needs significant effort
    • Decentralized implementation of biometrics requires significant organizational effort
    • Integration in different environments (single PCs, company networks) requires substantial effort by both, system providers (Biometric, Chip Card) and operators (system integration, administration)
    • Significant larger pilot implementations are needed (where the use of biometric systems is mandatory) before a massive implementation of biometric systems can be evaluated and recommended

[1]: Omnikey AG, [2]: Utimaco AG, [3]: Sparkassen Informatik, [4]: BioID, Dermalog, Ikendi, Iridian, Softpro, Wondernet/Dr. Fehr

 

Analysis of biometrics for potential use at automated teller machines

  • International and other complex regulations prevent effortless pilot implementations
  • Organizational and technical efforts are considered very high, as biometrics may only be used alternate to the PIN (liability aspects are more important than convenience offerings)
  • Consequent: It seems difficult to establish a business case for the next few years

 

Germany

  • Unique interdisciplinary project with over 30 project members, strong international interest
  • First comprehensive biometric survey of the general public
  • Providers/products from many countries: Germany, Austria, Sweden, Israel and U.S.A.
  • International conference participation: e.g. Barcelona, Berlin, Brussels, Budapest, The Hague, Frankfurt, Hamburg, Munic, New York, Qatar, San Francisco, Stuttgart
  • Unique skills building and improvements for project members
  • Contributions to international standardization: BioAPI, CBEFF (DIN, NIST, IEEE)
  • Numerous publications and books
  • Public workshops (Muenster 2000, Friedberg 2001, Berlin 2002)
  • Workshop presentations and additional information available at the BioTrusT website http://www.biotrust.de/