Page 61 - Automated Fingerprint Identification Systems (AFIS)

46 AUTOMATED FINGERPRINT IDENTIFICATION SYSTEMS

new plan was developed with two phases: an initial capability in 1998 and a full
operational capability in 1999.
With funding in place, a new organization being established, and a strong
commitment from Director Sessions, the Department of Justice, OMB, Con-
gress, and the criminal justice community, the challenge of specifying and
building IAFIS began in earnest. By May 1992, the new schedule was being
established and briefed. NIST started hosting workshops on the standards; Tom
Hopper was engaged in discussions and studies on compression techniques for
transmitting fingerprint images; and Tom Roberts and Walt Johanningsmeier
were doing the systems engineering required to specify the systems. Interest-
ingly, the three segment managers had already started developing their segment
requirements without benefit of a system level specification.
The decisions made in 1992 and 1993 still form the baseline for most AFIS
procurements around the world. The key decisions are discussed below.

2.7.1 TRANSMISSION STANDARD
Michael “Mike” McCabe of NIST facilitated a series of three successful work-
shops and produced a draft standard for the transmission of fingerprint images.
Given that there were competing livescan capture rates and that some AFIS
used binary images, the workshops required many compromises. The four bril-
liant facets of the standard developed are the following:

1. Each transmission has a header record (Type 1 record) that describes the
type of transaction (for instance, a miscellaneous applicant request or a
search response.) The Type 1 record also identifies the number and type of
records that follow. A Type 2 record, containing information about the
subject of a transaction, such as demographic and biographic data, or a
response, such as identification information or an error message, always
follows the Type 1 record.
2. There were four fingerprint image record types (Types 3, 4, 5, and 6) in the
original standard. Communities of interest could specify which ones they
would accept.
3. The fields in the records were tagged so that only mandatory and some
optional fields could be used without having to explicitly show all the other
optional fields as being empty.
4. Data fields could be specified as to their byte length, contents, and manda-
tory versus optional nature by domains of users. Among other uses, this
would permit Europeans and others to use the ISO standard format for date
fields while the FBI could use the American format.