Thursday, May 24, 2012

Personal fingerprint scanners seem Big-Brotherish, inevitable

Source: BBC
Part of me finds the idea of personalized fingerprint readers carried around by police officers abhorrent, even as part of me is already resigned to their inevitability. They're rolling them out now in the UK at Scotland Yard and the London Metropolitan Police.

But for what, precisely, would they be used? Judge Caprice Cosper in Houston has suggested repeatedly that deployment of these personal fingerprint scanners would make police more comfortable with using their cite-and-release authority for certain Class B misdemeanors because they could be sure at the scene who they were letting go. That's supposedly what's happening in England. According to BBC, "A suspect's fingerprint can be taken on the device and almost instantly checked against the police database. If a match is discovered, further relevant investigations can be made by an officer at the station. An NPIA spokesman said data from the scan is only used to check a match and is not retained."

Here's what I don't understand, and perhaps some informed reader can help me out: I've sat through presentations from Texas DPS fingerprint examiners who insisted that, unlike on TV (and apparently England) where fingerprints are matched in seconds by a computer, in Texas a real live human subjectively, laboriously compares prints for possible matches, and a second fingerprint examiner must verify it before a match can be declared. When a new set of fingerprints comes in, they go onto the stack until an actual person gets around to looking for a match, and that person generally has a backlog.

So if deployed in Texas, as Grits understands it these personal fingerprint examiners wouldn't let officers ID an individual on the street, they would only gather data which could be used later for whatever purpose. (In the UK, by contrast, the data "is not retained"). I can see where it might be useful for investigators interviewing suspects, witnesses, etc., for future reference. But as long as fingerprints are matched manually, these personal fingerprint scanners wouldn't let Texas police ID individuals at the scene. They'll still have to find out who they are the old fashioned way: By asking them.

20 comments:

  1. When you are printed and the prints are entered in AFIS, an AFIS # is assigned to attach those prints to your identity. Most states are able to identify you through use of this number, but I've also heard that explanation from DPS about certified fingerprint examiners pouring over scores of print cards.

    ReplyDelete
  2. "Here's what I don't understand, and perhaps some informed reader can help me out"

    It's an apples and oranges thing. The laborious process you are referring to is matching of latent prints, which are messy, smeary, incomplete things. As a rule. What the UK police are doing is using a mobile version of a fingerprint scanner like those used in security systems which generates a small digital fie that can be easily transmitted and searched against a large database. It's the biometric version of Shazam.

    ReplyDelete
  3. Add that to the 30,000 drones Nobama plans to have the USA before 2020. Nothing like a police state!

    ReplyDelete
  4. It's questionable whether police will feel comfortable enough with this technology to cite and release. While the technology is mature enough to check a fingerprint against police databases, there must already be a fingerprint record against which to match it. For example, a person previously arrested and printed should have an AFIS match. On the other hand, a person on whom there is an outstanding warrant for murder but who has never been arrested will not have a print record to match.

    No system is perfect, but this technology moves the ball forward just a bit.

    ReplyDelete
  5. @ Thomas Hobbes...very well thought out and "right on spot" comment.

    ReplyDelete
  6. But Thomas, they can't arrest every Class C offender on the off chance they might be lying about their identity, and they could be murder suspects, too. On the list of petty Class Bs they're allowed to cite for - MJ possession, no insurance, etc. - I don't think one can justify prospective detention based on the distant possibility somebody might be a murder suspect.

    ReplyDelete
  7. Oh, and thanks to Texas PO and 9:35 for the explanations. That makes some sense of it.

    ReplyDelete
  8. I always thought this was why DPS required a thumbprint when you obtain or renew your driver's license. I know some agencies use similar print readers to obtain a preliminary identity on deceased persons with no identification, but maybe that's all DPS is using those thumbprints for. But this really does feel like Big Brother reaching out to slap us all.

    ReplyDelete
  9. Pretty clearly it's a bit of a "damned if you do, damned if you don't" situation for law enforcemnet, isn't it? If they use the devices they're declared to be "big brother"ish. If they don't use the technology available to them, and then inadvertantly release someone who's got a warrant out on a violent felony, they're going to catch some serious criticism for that, too, right? Perosnally, I'm fine with them using the fingerpriant scanner, but I think they ought to have a policy about not dragging people in for unpaid parking tickets if the scan shows a warrant out...

    ReplyDelete
  10. Why is Houston such a dupe for all of these Big Brother programs? We've seen: DHS trolling busses, drone deployment before the FAA even legally approved police to fly them and now this? What is going on at Houston City Hall that they believe they can just pass this off onto Texans? I'll tell you this, locals have better use their voice before they lose their voice if they don't want this on their streets. Local police should only get your prints if you are arrested, booked and in a jail cell. This makes it way too convenient for them to treat non-arrestees as criminals. This is not okay. Houston, you do not have to put up with this.

    ReplyDelete
  11. Grits, you're acting like I endorsed the failure to use cite and release. Personally, having written my share of traffic tickets, I don't know why departments are so resistant on low-level, nonviolent misdemeanors. It's not always necessary to arrest someone to allow the system to address their offense.

    I simply wanted to point out that the marketing (as often happens with CJ applied technology) may have inflated expectation beyond capability. You can't always be sure of who you're releasing.

    ReplyDelete
  12. The DPS article you refer to, is it not, in relation to latent prints lifted at crime scenes for comparison into the AFIS system?

    LIVE SCAN submits ten print impressions during an actual booking in relation to the CJIS requirements and are compared against ten print impressions recorded in AFIS.

    The use of this technology would have great value in the field, especially as it relates to a known perp trying to use your ID. Imagine the pain in the ass ordeal you would have to go through when some crook uses your ID during a stop and the officer has no way of knowing for sure if the ID is valid.

    You don't show up for court because it wasn't you who got the misdemeanor field citation instead of being arrested. The perp doesn't show up for court and now a warrant, with your identifying information, is generated and entered into the statewide warrant data base TLETS. You get stopped for speeding, the officer queries your name in the system and now you get arrested on the warrant because the real perp's fingerprints were not taken in the field at release. Or maybe you don't get arrested because the officer had the technology available to take the prints in the field and can now determine it was not you who got the original misdemeanor field citation.

    ReplyDelete
    Replies
    1. LE can use a mag stripe reader or barcode scanner to retrieve your information printed on the ID or DL. That picture on the front had better match the one on file with DPS.

      The scary thing about taking finger prints in the field is getting them mixed up - accidental or otherwise.

      Delete
  13. I could see someone use this to wiggle out of carrying ID or even a DL. If they dont trust photo ID, why bother.

    ReplyDelete
  14. "A Texas PO" is partially correct. Yes, an AFIS number is attached to a person's prints entered into the system. But this is only an internal departmental "tracking" number and not a "matching" number. It is used only when that particular set of prints recorded in AFIS's databank is needed to be utilized again. AFIS results are not supposed to be used "as a match" like you see on TV or they suggest. AFIS will run an algorithm based on the 1st level detail (the Galton details- formerly known as "points").

    A fingerprint expert then takes the results from AFIS, which lists the top ten closest (or whatever number they have it set for)"hits" with a percentage of how close the match is. It also shows the FBI number, the local PD's numbers (Because it is their system), CII number, or the NCIC number. The examiner then pulls the tenprint card on file using one of the above reference numbers to use as an "exemplar". The name should never be looked at. The examiner then compares the exemplar to the latent prints. An AFIS printout will be used in lieu of a tenprint card if there is not one available. Cards can also be ordered from DOJ and the FBI. The examiner than uses the ACE-V methodology (Hopefully not the old "point" system) to conduct a side-by-side comparison of the available exemplar to the latents. Going from Galton detail to Galton detail until every last detail has been checked. This will also take into consideration of 2nd and 3rd level detail- which is the positive/negative spacial relationships, the pores, hair follicles, etc. After the comparison is completed, a second and even sometimes a third verifier conduct their own independent comparisons. AFIS is not supposed to ever be used for the sole source of "a hit"- only a starting point.

    Now, if you have one of these biometric systems, it will scan the prints and upload them into AFIS. AFIS could even kick back a percentage rate of what the "match ratio" might be right then and there to the officer. This is where every qualified fingerprint examiner that knows their stuff cringes. This "hit" can be wrong.

    We have learned with the antiquated "point" system, there have been so-called "matches" based on 16 points that aren't even the same pattern type- a loop and a whirl.

    Even though AFIS is extremely accurate with it's matches, it's are still basically working on the point system when creating it's algorithms. AFIS is only to speed up the process by narrowing down the search. All AFIS results must be checked by a trained, competent fingerprint person.

    Two different fingerprint patterns can sometimes be almost identical in every way, but if you look close enough, you will find a difference. The "no two snow flakes or tree leaves are ever perfectly alike" theory holds true.

    ReplyDelete
  15. Part 2: (ran out of permitted characters)
    During the 6th through 13th week of fetal development, the fingerprints are forming. A fingerprint is made up of single cellular units that through differential growth rates may or may not "connect" with another cell unit and create various details of the pattern: islands, dots, ridges, bifurcations, etc. Mono-zygotic twins have identical DNA but their fingerprints are different because of differential growth rate. The prints may be so similar that it takes a very close examination with a qualified examiner to locate the difference.

    As for latents versus live scan print (biometric)- latent impressions often are much more clear than live scan. Flexar skin is playable- this can cause "explainable differences" but an expert has to be able to use deductive reasoning to decipher. AFIS does not have this capability. Or, if the screen isn't clean, the scanned image (of the print) can become an "overlay" with another print. Or even a "double tap" might occur. Too much pressure can cause the matrix to blur or cause a reversal. Scanned prints can also be blurry.

    And all fingerprint impressions, whether latent or biometric, are "incomplete". A fingerprint is 3 dimensional. A finger is not flat. Even when you "roll" a print from side to side to obtain an inked print, if it is not perfect showing the entire detail area which stretches to the sides of the finger and onto the tip. A very high quality rolled print is close, but not "complete". This is a widely held misconception.

    And alas, someone does have to already be "in the system". Police departments are not nationally connected. There is NCIC, but that is "their" system. The prints are not universally "shared". Some systems are shared with a couple of departments and DOJ, etc. but these are arrangements.

    OK. Did I miss anything? Then I’m off to bed.

    ReplyDelete
  16. I did miss something. Anon 17:1300 mentioned that the field live scan scan could eliminate someone if they use your ID. Yes, this could work. Safer to use for "elimination" than "match".

    ReplyDelete
  17. The Fishing Physicist5/30/2012 11:52:00 PM

    Such devices are just like red light cameras. A fraud being sold to government by con-men to make a fast buck.

    Two words: Junk Science

    ReplyDelete
  18. Technology like this and the attitudes underlying it are, in large part, why I left the UK and settled in Texas. I would be very disappointed to see this great state go the same way, with the lies and half truths as justification.

    ReplyDelete
  19. There are plenty of reasons these scanners are bad, but I'm just going to throw out one thing here.

    Much to my surprise, it seems Texas law protects people's right to refuse to identify themselves to the police (I don't think this applies if you are driving).

    The exact wording of the statute is at the end of this comment, but the summary is that there is no penalty (that I can find) for refusing to give any personal info to police. The exception is if you are under "lawful arrest". Even then, the penalty is a Class C misdemeanor (I, personally, would take the ticket).

    It seems these scanners could both take away our apparent right to refuse to identify ourselves and I also wonder if police randomly using these scanners is a violation of the 4th amendment. I haven't researched or even thought it through, so I don't know, but it feels like it.

    PENAL CODE Sec. 38.02. FAILURE TO IDENTIFY. (a) A person commits an offense if he intentionally refuses to give his name, residence address, or date of birth to a peace officer who has lawfully arrested the person and requested the information.

    (b) A person commits an offense if he intentionally gives a false or fictitious name, residence address, or date of birth to a peace officer who has:

    (1) lawfully arrested the person;

    (2) lawfully detained the person; or

    (3) requested the information from a person that the peace officer has good cause to believe is a witness to a criminal offense.

    ReplyDelete