Confirming AFIS hits, ASCLD considerations

[Alicia Wilcox]

I am looking for some feedback on what other agencies are doing when they get a hit on AFIS. Up until now we have always confirmed the identification on the screen and then driven to our repository and requested a physical card. We then confirm the identification using the card. Sometimes the card is an inked card, sometime it’s a Live Scan (electronic card) print out.

We are ASCLD-LAB accredited and are wondering if there would be issues with us making the ID on the screen and calling it good. Would the images in AFIS then be considered evidence and need to be tracked as such?

Another consideration is the possibility of images in AFIS being associated with a wrong SID and then a wrong name. If they give us a live scan print out to confirm our ID there is also the possibility of a wrong name being associated with the fingerprint images. Sometimes only a live scan card is available, as we go more towards electronic cards and less inked cards.

What do other labs do? Are there special ASCLD considerations?

Alicia - Maine

[Pat A. Wertheim]

Hi Alicia

Our lab is ASCLD accredited. When we get a "hit," we enter it on the page as "null" and complete the comparisons with the rest of the candicate list, then "pause" the transaction. After we have obtained the 10-print card, made the identification, and had it verified, we go back in and declare it a "hit." None are declared "hits" unless they have first been verified. Although it has never happened in my lab, if a "hit" were excluded instead of identified, we would go back in and declare it "no ident" and close out the transaction to put the latent into the unidentified latent print file.

At first, I had my doubts about this procedure, but it has worked very well.

[Steve Everist]

Alicia,

Are you asking if there are agencies that have taken the use of 10-Print cards completely out of the equation and are only doing comparisons resulting from AFIS candidates from the on-screen images?

If so, I think Pat answered a different question than you asked.

If that is the question you're asking, I'm also interested in the answer. Relating to another recent question, we still do our final comparisons using the 10-Print card. But I'm guessing there are agencies that are doing some of this electronically.

[sharon cook]

My agency is ASCLD certified. We have AFIS Techs who enter and search the latent prints. When they feel they have a hit, the information and side-by-side documents are printed out, a request for comparison is done by the Techs and then the whole case is forwarded to a Latent Print Examiner to complete. An AFIS hit is an inquiry, not an analysis...so far. An AFIS "hit" provides probable cause for an LPE to focus the search on one individual. It is still the investigatory stage. We still have to either get the inked card or the electronic printout and make the comparison to determine if it is an identification.

We consider the AFIS printouts to be part of the case documentation, so they are attached to the case in our LIMS. If there is wrong information on the printouts, we correct it by hand, and date and initial the changes.

Feeling that the AFIS candidate is a hit is by no means final. Nothing is final until that official report goes "out the door."

[Alicia Wilcox]

Thanks for the feed back.

Here in Maine the latent print examiners are the "AFIS Techs" as well!

I would like to clear up some confusion. We do not let the computer make a hit. When I said AFIS hit I meant the LPE determines its a hit to a candidate that AFIS returned.

Sharon, you say you still get an inked card or electronic print out to confirm the identification, that’s what we do here too. But why to we need an electronic print out of the screen when we can see is an ident on the screen. If the resolution was bad or if the LPE couldn't confirm the ident, by all means a card is necessary.

What I am wondering is if we can go 100% electronic when we have an identification that is confirmed and verified/tech reviewed (on the screen and with images printed out from AFIS) with out the inked card or electronic card?

Alicia
[/quote]

[sharon cook]

Alicia; AFIS printouts are documentation. ASCLD requires documentation up the wazoo!

Identifications must be made using known exemplars...not the AFIS screen. Most of the time, the split screen images are so poor, we couldn't use them if we wanted to, ayway.

Besides, what if there are other latent prints in the case? You need the exemplars to compare everything in the case.

[RL Tavernaro]

I think Sharon makes a good point regarding the need to compare all the latents to that subject (assuming an ID is made to the original AFIS latent). Although some AFIS systems may allow on-screen comparisons, it usually too cumbersome to be a viable alternative. An actual tenprint card for manual comparisons typically will be much more efficient.

In addition, even if the AFIS latent is the only latent print in the case, what would you provide to the defense under discovery, and what do you submit to the court under 'best evidence' rules?
Regards, RLT

[Brigid Lockhart]

To those with ASCLD or ISO experience:

Our lab is attempting to gain accreditation and our boss raised the question of confirming suspect's identity based solely on information in the AFIS archive.

Here's how we do it now; enter a latent, compare list of candidates on screen, if a possible suspect is developed, the SID# is written down and then that SID# is queried in the "AFIS archive" which is the database of tenprint cards/livescans maintained by our state. The tenprint card is printed and we compare that card to our latent to confirm an ID (verifications and etc. still follow). The question my boss has is if the archive we print out is sufficient for pinpointing a particular person, or if we cannot really do that until we look at an INK card from the person when they are in custody.

Secondly, if that card is printed from the archive database, do we treat it as "evidence" and lock it up/seal it with the latents or is it enough to have it as a page in our report as case documentation.

I believe getting an ink card from every arrest is not feasible and it is cumbersome when you have to repeat your exam a second time. If we cannot rely on our state-maintained, secure archive, then we really can't be sure of anything. The subjects may lie about their names at the time of arrest, but I can identify the same impressions on 10 cards with 10 names, and one SID. The detective should be doing the person's identity verification.

Please share your insights!

Thanks!

[sharon cook]

Dear Brigid--good luck getting an inked card for every ID! Inked cards are few and far between these days. In our agency, any tenprint or palmprint exemplar printed out from the livescan database is an original document.

We do not consider a tenprint or palmprint exemplar taken during the "normal course of business," (during booking), to be evidence, requiring packaging and seals, etc. They are business records or reference materials. When we do print out an exemplar and use it to make our identifications, we keep that copy in a file we have set up for that, or we scan it into the case in our LIMS.

[Brigid Lockhart]

Sharon,

Thanks very much-that all makes perfect sense to me!

[mdavis]

I'm a bit late here on perhaps a dead topic, but I'll throw this in the pile.

Our AFIS "hits" (I don't like that word) are candidates only because our AFIS people are not examiners. Once a candidate is individualized (I don't like that word either ... see multiple other postings), then we use a flatbed scanner or digital photo of the print and the control. A scanned print at 600 or 1000 dpi and printed on a 600dpi laserjet printer is quick, inexpensive and adequate casefile documentation in almost every case.

[reidsm]

My lab got accredited by ASCLD/LAB last year.

Once we get a "hit" from AFIS, we still get a copy of an actual fingerprint card prior to officially making the identification of a latent print.

I think the most important thing as far as ASCLD is to document everything. In your policies you need to say what you do and do what you say.

Sharon