AFIS Suitability

[Charles Parker]

State police have also refused to turn over lab notes on the fingerprint, and they won’t make public their manual that says which types of fingerprints are suitable for comparison to the AFIS database and which ones are not. The manual is a trade secret, state police have told Bill Clutter, investigator for the Innocence Project.

However, the Internet is rife with information on AFIS and what fingerprints can be compared to the database. Via Google, Clutter has found an online Cincinnati police fingerprint standards manual — the same type of document Illinois State Police want to keep secret — that says how much of a person’s finger must be captured by a print before the AFIS database is useful.

I have never really seen a manual which dictates what types of fingerprints are suitable for AFIS. In talking with other agencies there seems to be a wide variety. I have been told by some agencies that they need a core or delta area (sometimes both) before they will run one in AFIS. I have seen some very poor latent prints that had only flowing ridges with no core or delta identified through an AFIS Inquiry (of course I have seen a lot that were not ID either).

If an agency had a standard of suitability for an AFIS Inquiry is that standard applied equally to a latent print from the exterior of a vehicle in a "Burglary of a Vehicle" as it would from a "Homicide Scene".

Are latent prints so static and uniform that a standard could be developed on what will be searched and what will not-----what about the differences between two examiners judging those "borderline" latent prints.

In my own office I have seen latent prints that were not run in AFIS that I thought perhaps taking a shot at it might not be a bad idea. I have also passed upon latent prints that I did not think were AFIS suitable and another LPE thought "why not?".

Is there any uniform training in what is AFIS Suitable and what is not? You have a number of different vendors out there-------are their standards as to what is suitable for an AFIS Inquiry and what is not DIFFERENT?

I wonder what they mean by determining what is AFIS Suitable and what is not a TRADE SECRET? I realize the searching program for different vendors is proprietary but determing what is suitable and what is not for inquiry-------a trade secret.

[Bill Schade]

If you can identify it, you can plot minutia and an AFIS system can return good results.
"AFIS Quality" is a myth that began with systems in the 80's.
The only standards today might be which unidentified latents are left behind in the database for "reverse searching"

[Charles Parker]

Bill, you hit the nail on the head. I have seen some real crap hit in AFIS and the comment about what to retain for reverse searches is key statement. You might want to run it but not retain it in the unsolved latent print data base. A recent book on AFIS broaches that question.

A third element is the number of people that are trained and are is their personnel available sufficient to inquire a number of poor quality latent prints for AFIS Searches. The manpower requirement; is it sufficient.

[Pat A. Wertheim]

Our AFIS requires a minimum of 3 characteristics. Three or more characteristics in a latent make it, by definition, suitable for AFIS. Now, who on AFIS today would run every latent print with only three points? Nobody that I know of! So if the latent has three or more points and I chose to NOT run it, am I ignoring suitable latents? By definition, again, yes, I am not running AFIS-suitable latents.

But I have seen too many high score #1 candidates with 8, 9, 10, or more points that are exclusions, based on other points outside that 8, 9, or 10 point cluster. I saw one the other day a friend emailed me that conceivably you could take to 19 points. But it didn't pass the "feel good" test, as Ron Smith used to call it. (Please do not ask to see that print. I was sent it in confidence and will not pass it on.) So the question becomes, at what number of points do I personally feel comfortable entering a latent into AFIS? Several considerations: Does it have enough points (subjective) to avoid too-close hits? Is it free enough of distortion to be sure of what I am seeing? Does it have some good Level 3 detail to add weight to an identification?

One examiner I know proposed a written policy a few years ago that I thought was reasonable, but his suggestion never got off the ground. I don't remember the exact wording he proposed, but it included both Level 1 and Level 2 minimum requirements. So, for now, three points means a latent print is AFIS quality on my system. Anything with three or more points that I do not run means I am ignoring AFIS quality latents.

That puts us all in a tough spot in the witness box, especially when there are unidentified latent prints and the case against the defendant is circumstantial. But it's something we have to live with, I guess. Bottom line: in my agency, the decision whether to search a latent through AFIS is completely subjective (as long as it has 3 or more points).

[Bill Schade]

I knew when I posted that this thread would get to this discussion eventually.

Pat,
Your point is well taken but I don't think the issue in court is any different than when a crime scene technician has to explain why he did what he did at the scene. Which processing technique he (she) used and usually which ones he (she) didn't use.

There is a lot more to it than "how many points" regarding AFIS searches. I think that is the problem with written policy sometimes. It's difficult to cover every possibility and can thwart initiative.

I do not question every decision my staff makes in working their cases, They know that the lawyers (and sometimes investigators) will do that.

And if you want to really start a discussion, what about when a print is run in AFIS, the subject turns out to be in the database, but a hit was NOT made via the AFIS search.

Is that a mistake? Perhaps a false "exclusion" (exclusion being another topic we could debate over an adult beverage)

Simply put, I'm amazed all the time by the prints my staff identifies, AFIS searches are a big part of our arsenal these days and anyone that does not utilize it when it is available to them is not getting their moneys worth out of the investment in technology..

Just my personal opinion.
Bill

[Charles Parker]

I do not know how low we can go to enter a latent print in our AFIS, but the lowest I have gone is 6 and I did a 7 one time and got a hit. Technically I can run “frog lips” in our AFIS but they are not going to hit anything (at least I do not think they will). Besides the amount of 2LD for me to enter one I am looking at the clarity and reliability. I do not remember considering if 3LD is present as a factor to entering one in AFIS. If it is present I will use it. If not but I have clear and reliable 2LD then I will probably search it.

At one time I use to register into the ULDB everything I searched. I have since stopped doing that and probably only register about 70% of what I search. Does that mean I am not taking advantage of reverse searches? It is an experience call----nothing more – nothing less.

Bill, I agree with your posting. There is a lot more to AFIS than just “how many points” such as clarity, reliability, distortion and from my POV case status which may not fit into some other people’s concept.

Like you I think for a court issue it is no different than other decision making/problem solving aspects of latent print examinations. I would add that we need to teach/instruct the courts and investigators about what AFIS can and cannot do.

I love the word initiative and that written policies can dampen initiative if not written properly. But then again I wonder why some LPE have initiative and some do not?

Your issue about when a latent print is not a hit in AFIS and the subject is identified later and his prints were already on file. We have had several of those and it has taken awhile to dispel from detectives the old myth that if it does not hit in AFIS then the suspect is not in there. They did not understand that AFIS is not 100% because that is what they were told 10 years ago and why has it now changed. “You mean the other LPE were wrong”---‘Yep---they were’.

I thought I read a story a few years back where somewhere in the northeast a latent print from a murder scene was entered into AFIS and it did not hit. Several months later the subject was identified with the latent prints that were run in AFIS and there were accusations from the detectives about the sloppy fingerprint work. If I remember correctly they were going to can or get rid of some of the ID people because it was believed that the miss was attributed to sloppy work and not because AFIS did not hit the subject.

Exclusionary searches by AFIS with later individualizations can be a nightmare if everyone does not understand the process. A good topic to discuss over some cold ones.

[Bill Schade]

Charles

Regarding reverse searches, We too have limited what we retain in the unsolved latent database and our "hits" from reverse searches went up dramatically.
Reverse search verifications are a whole other kettle of fish from latent searches against the ten print file. Totally different technique and in the case of my agency, done by ten print technicians, not latent examiners. We have outstanding results with this arrangement for a number of reasons.

So, I think you ARE taking advantage of reverse searches by limiting what you retain.