Page 263 - Automated Fingerprint Identification Systems (AFIS)

248 AUTOMATED FINGERPRINT IDENTIFICATION SYSTEMS

Table 10.1
Unsolved Latent File Workstation Coder SAFIS Matcher TPlc Database
Matrix
Original Original Original coder conversion
New New Original coder conversion
New New New coder conversion

burglary), and a priority. The first of the two lists contained the cases that had
been entered when the new coders and matchers were installed, but when the
search was against the database coded with the earlier version of coders. This
was akin to the best-quality latent having been searched against a less than state-
of-the-art database at the time of the original latent print search on SAFIS. The
second list contained the cases entered using the original coders, matchers, and
converted TPlc database. Re-searching these cases from the UL file (referred
to as reactivation) would require slightly more effort.
A series of training sessions was conducted at each latent print site in the
state that used SAFIS. The systemic upgrade of SAFIS was explained, as was the
immediate and potential benefits of the improved coders, matchers, and re-
coded database. The preparation and use of each list was reviewed, and the
potential for making more identifications through relaunching existing cases
on the UL file was highlighted. Since the lists provided the crime type for each
case on the UL file, examiners could quickly identify high-profile cases, such
as homicides, and search those first.
The latent print examiners were encouraged to relaunch the cases on the
first list and search them against the improved database, a relatively simple
process. Since these old cases were now being searched against a better data-
base with better placed minutiae, the examiners were able to make identifi-
cations. Many of these identifications were from records that existed in the
database at the time of the original search, but that, due to the limited tech-
nology at the time, did not receive a score high enough to appear on the can-
didate list.
The results of this effort can be seen in Table 10.2: the number of annual
latent print identifications doubled over the time period from 1998 to 2001.
The benefits of this overall system upgrade, combined with the manage-
ment decision to exploit these opportunities, produced amazing results. These
two elements provided the tools for the latent print examiners to do what
they do best: examine candidate, make comparisons, and ultimately make
identifications.
These improvements were carried out over several years. The introduction
of new workstations in the fall of 1998 was followed by the introduction of new