by clpexco » Mon Aug 10, 2009 3:55 pm
FROM THE REALLY OLD GUY AND HIS CORNER
By Richard Reneau, CLPE
FROM THE REALLY OLD GUY AND HIS CORNER:
This article is from one of the “old guys” currently involved in identification (or is it individualization) of fingerprints (and palms and plantar surface of the feet, to be accurate). It is not meant to be a scholarly treatise of the science or even a wise, solemn, assessment of where the science is. It is from a person who over the last 39+ years has been practicing the profession -- from a person who has lived it not just studied the historical record.
The following is not the message but just a recap of the centuries before I became involved -- the message will come briefly at the end. For a first hand report of the centuries before me, please contact others still practicing in the field.
ACKNOWLEDGEMENT:
The realization that man(kind) has patterned ridges on his hand pre-dates the Christian era by many centuries. As time passed, developing civilizations noted these patterned ridges in various ways. For instance, on the face of a cliff in Nova Scotia an illustration exists of a hand showing crude representations of friction ridges. The Chinese also recognized the value of friction ridges on the fingers; documents in the eight century of the T’ang dynasty note that fingerprints were impressed on business contracts.
SCHOLARY WONDERMENT AND THE SCIENCE OF FINGEPRINTS:
The arguable start of the science of fingerprints was around 1684 when Nehemia Grew, a fellow of the Royal College of Physicians, commented on fingerprints during the course of lecturing.
Moving to a much later period, many other identification methods were being developed also as the science of fingerprints emerged. Most notably are the Bertillon system of anthropological body measurement and facial photographs. Each of these methods was employed on a limited basis until circa 1904 in Leavenworth, Kansas. There the famous case of William West took place. At that time the fallibility of the photograph, the anthropological measurement system (Bertillon method) and name matches were discovered. The event itself showed the emerging profession something, but as usual it took a while to realize the real significance.
No surprises yet!
AUTOMATION:
Finally, law enforcement, specifically the fingerprint profession, progressed into the space age. Considering the wide-spread application of Electronic Data Processing (EDP), it was only a matter of time until the technology could be applied to dactylography.
Much of the beginnings on automating fingerprints were accomplished by the FBI. In 1963, the FBI consulted the National Bureau of Standards (NBS – now the National Institute of standards and technology or NIST). The study concluded that automation was feasible, and further provided an analysis of the current technologies available. The three technologies available at that time were:
(1) Continued to allow fingerprint technicians to classify, then place the classes into a computer for future use.
(2) Optical correlation by overlapping two prints and comparing global characteristics.
(3) Digital imaging by scanning the print and extracting identifying characteristics.
As I recall (from talking to the really “old Guys” in the coffee shop, of course), and as an aside, the FBI had to make this an ongoing project for a while until the technology could mature into what was needed.
The first choice was rejected because the process still remained labor intensive and would have resulted in even more that the 1400 fingerprint technicians already employed. The second choice was reviewed only to find that optical correlation did not have sufficient selectiveness to be feasible for the large number of existing records.
The FBI then narrowed the choices to digital imaging developments. In 1967, contracts were awarded to Coronell Aeronautical Laboratories Inc. (which became Calspan Corp. and later, wasn’t heard from much) and Autonetics Group of Rockwell International Corp. (which was sold to Dalarue of London, sold to a management group, sold to Motorola, sold to Morpho) for engineering. Significant results were obtained by 1969 and further developments proceeded (development costs from 1971-1977 were a mere $21,535,700, by current spending standards a real bargain).
It is important to note that much of the original FBI research was done with the idea of using the technology for classifying and searching ten-print “cards.” For in this way, the FBI could reduce the tremendous labor cost associated with manual processing (and possibly improve accuracy). Also important is the fact that the research proved successful and the FBI even today still uses these systems (currently, more appropriately, grandchildren of these systems), with a corresponding labor force reduction to approximately 300 (I originally wrote this circa 1987, today the numbers might be slightly different).
The first prototype fingerprint system developed was called “FINDER.” A system I remember viewing in FBI DIVISION #1, when I worked there. The future at that time was rather “clunky” by today’s standards and utilized a concept of a “flying Spot Scanner” to extract minutiae. The “flying spot scanner” used a narrow beam of light to scan a fingerprint one at a time and note variations of the print (black) against the background (white) of the card, thus giving rise to the ability of detecting an ending ridge, etc. It is noteworthy that the detection of minutiae happened on a fingerprint card and not by a picture in computer memory which is “processed.” I remember seeing the process represented as a movie which was produced to show all FBIers what the future looked like (maybe we can get a copy from the archives and re-master it for show and tell). Also, the profession started to get neat (also see groovy) words like minutiae in its everyday vocabulary.
[NOTE: Much of the previous information from “Acknowledgement of Ridges” to this point was gleaned from federal publications, published articles and publications in which I was interviewed and now must quote someone else as the originator. Such is life. In short, I have misplaced the citations.]
AUTOMATION GAINS TRACTION:
This is where my memory gets much better.
Emerging with the ten-print technology was another idea - The idea of being able to take a latent found at the crime scene, search it, and identify it out of an electronic data base. This idea was researched by the California Department of Justice (DOJ) and the San Jose Police Department. Both agencies received an LEAA (Law Enforcement Assistance Administration) grant to research the possibilities. Each of these agencies received a prototype system from Rockwell International. Later, California, DOJ, purchased a system which they utilized until the advent of the State CAL-ID program. On the other hand, San Jose experienced considerable contract problems and in October 1978 returned the system and entered into litigation against Rockwell International (I know I was there). The settlement resulted in San Jose obtaining a Printrak 250S system from DeLarue Printrak.
With the success of the FBI and its partners, other companies soon began developing/marketing systems for use in both of the newly defined areas of automated fingerprint identification. Many of the smaller agencies, by comparison these agencies could be very large but with a smaller number of employees than the FBI, could not get the same cost reduction that the FBI enjoyed for the ten-print process. So many prospective vendors introduced the concept of the automated fingerprint systems performing the “Big Four.” The big four functions were:
(1) Current arrest ten-print to a known ten-print data base.
(2) Current crime scene latent prints to a known ten-print data base (prototypes appeared around 1977).
(3) Unknown latent to unknown latent searches.
(4) Current arrest ten-prints to a previously unsolved latent data base.
As the concept of Automation started gaining traction, a spate of reports/studies etc. happened, first from California State (because that’s where I am, or at least I think I am still in CA), then federally:
“CAL-ID Remote Access Network Study” by Applied Systems, for the state of CA, indicating work load and system placements. (I had to call another really old guy to refresh my memory on this one.)
Direct Image Retrieval Report, “DIRS Subcommittee” of the CAL-ID Operational Advisory Committee;
The “Status Report: CAL-Identification System and Remote Access Network (RAN) for Calendar year 1987” (Second year of a five year report mandated by the state legislature on the state of fingerprint automation, after purchase);
The “Increasing Efficiency in the Criminal Justice System: The Use of New Technology for the Criminal Identification and Latent print Processing,” also known as the RAND report or the Ratkovic report (this report is cited as the basis of fingerprint automation within California);
“The FBI Fingerprint Identification Automation Program: Issues and Options, Background Paper, By the Office of Technology Assessment Congressional Board of the 102nd Congress (November 1991), ISBN 0-16-035974-0.
The ANSI/NBS-ICST 1-1986 (and Prior and later) “Data format for the Interchange of Fingerprint Information” for ANSI/NIST standard.
[NOTE: These are the ones I remember and participated in, maybe you can recall others. These reports signify the birth of automation and the following accountability after implementation.]
This approach by vendors to the emerging market and the studies, provided managers and the profession with a methodology to perform cost benefit analysis to justify the, then, high cost of start-up. In California a study which was utilized by many was the “RAND” report (AKA the Ratkovic report – yea, I was around for that one). Cost benefit was established to the satisfaction of the legislature and the Attorney General to the point of purchasing a state-wide Automated Fingerprint Identification System application. Along with the state’s purchase, the requisite political sausage produced by cities and counties resulted in funds providing for the initial purchase of many other stand alone systems throughout California.
AUTOMATION SEARCHES FOR SIGNIFICANCE BEYOND COST BENEFIT:
Circa 1985 automation was birthed, at least in California.
What is very interesting to note from the old guy’s perspective is that much of the early “sales” of the Automated Fingerprint Identification System (AFIS) centered around “hits” made from searching latent prints in the system. AFIS capabilities as a crime fighting tool quickly overshadowed its speed and accuracy of the ten-print searching and identification abilities. The latent print function of the “Big four” quickly captured the public’s attention on the news. For example, one of these political sausages was the direct result of the automated systems “hit” of the “Night stalker” case. It was all about getting your own stand alone system at that point.
A central discussion I recall at the time was whether we should let the crime fighting capability take the lead -- won’t wide-spread publicity cause the criminal to stop leaving prints at the crime scenes. Since the release of fingerprint information hasn’t stopped the criminal in the last 100 years, it probably won’t stop now, we opined. How right we were --backlogs are still growing. But I digress – back to the story.
MEANWHILE BACK AT THE RANCH -- THE NATIONAL SCENE:
A curious situation was developing in which the profession would not realize the true extent of the change for many years (sound familiar, just like the William West). I recall speaking with Bob Hazen around 1984 about the changes now being dictated by management, Congress and a weak profession not able to protect itself.
To wit, the reduction of the FBI work force from 1,400 to 300, the changing job function and the problem associated with finding the previous cream-of-the-crop-highly-qualified fingerprint professionals ready to promote to latent print examination. Back in the early 1970s, many in the FBI could not apply to latent prints without first having completed 5 years in classifying and searching – 100’s to 1,000s to tens of thousands to hundreds of thousands of comparisons of fingerprints, day in and day out – whether it was the day after the night before or not you had to perform.
Meanwhile, congressional budget allocations forced the FBI to change the location of fingerprint operations and ultimately split the latent print processes from the ten-print process. The separation was miles and states apart. Also, many states and localities eventually followed suit.
So what! As we look back with objective professionalism we might note that the physical bifurcation started manifesting as a divergence of management and professional philosophy.
ZEAL FOR AUTOMATION MORPHS ATTITUDES:
Within the profession, we subconsciously considered the ten-print personnel as being somewhat less than the “real professional” – the latent print examiner. This philosophy has, in my opinion, given rise to the concept of ten-print examination being a “clerical” function. And has taken root in the different management styles and demands of the ten-print examiner verses the latent print examiner.
It has been stated that the difference between the crime scene investigator and the lab technician (who just develops or photographs) is that the “experts” can and do make a comparison of the fingerprints (friction ridge detail), then form an opinion and are therefore allowed by the courts to render that opinion.
However, the ten-print personnel render opinions every day many, many times about the identification of a person. Couple this with the concept of the “clerical” function and the outcome can be seen as applying a “best-eight-out-of-ten” approach to personnel. In this environment the ten-print personnel can be seen as getting the last eight out of ten correct answers – which is 80% -- a solid “B” average – this is good (a little editorial license used here).
In the latent print environment, we expect a near 0% error rate or the person may never be allowed to testify, even with just one mistake committed over decades of experience. It seems, however, that the disconnect between the fingerprint identification process of the ten-print examiner and the fingerprint identification process utilized by the latent print examiner is causing confusion about the whole process of identification.
There is, of course, a difference between what the ten-print person does and the latent print examiner. But there still can be a situation set where a latent print lifted exhibits the same quality and quantity as a rolled impression. In this situation any missed identification or erroneous identification would be viewed with different criteria, depending on whether the person was a latent print examiner or ten-print examiner.
Consistent treatment of the fingerprint identification process across the spectrum can be used to moderate the confusion between the fingerprint identification processes practiced by the different job functions.
Another very interesting concept is that the court (or at least the defense attorneys) has used ten-print errors committed by latent print personnel as part of the “individual error rate.” I think the lines have definitely blurred sufficiently to require movement in addressing this issue.
Just looking around the “old Guy” is encouraged that SWGFAST is trying to integrate the identification, training and documentation processes for both ten-print and latent print comparison.
THE PESKY BRIEF MESSAGE:
At last the promised message which appears briefly at the end is here.
How many Daubet hearing have there been regarding the science of fingerprints and the practice thereof?
A more interesting question, in this “old guy’s” observation, is how many Daubert court challenges regarding the science of fingerprints are based on the professions inability or ability to identify a person in a ten-print environment?
The history lesson at the beginning and the progression to the science, I hope, shows that there was never a conscious effort to consider a difference between the identification process utilized by the ten-print personnel and the latent print examiners.
The profession itself has created two different certification processes administered by two different boards. Has the profession bought into the same divergence of function process? Has this allowed the academic community to “cut the latent print processes from the heard” so to speak, so that print identification can be more effectively challenged with the “divide and conquer” approach?
When automation is considered, it is important to note that the automated systems and the matching algorithms were extracted from the heads of many of us really “old guys” still practicing. And the truth be known, automation is responsible for a good guestimate of about 75% of the arrest and applicant identifications in the US, without a person being involved (Known as “lights-out” arrest processing) – until testimony or the identification is called into question.
The millions of identifications effected for those arrestees and applicants each year and the tens of thousands crime scene identifications effected each years are the same thing – the ability to effectively identify someone through the friction ridge comparison process.
This is not to say that basic research is not needed or would not be profitable; but the latent print community is not alone – there is an industry of computer companies, ten-print examiners and latent print examiners, who are all in this together.
The clock is ticking; but that is another story.
