Using License Plate Readers to Fight Crime, Part 2 of 4, Meghann Tracey, Project Manager, Technology Center, International Association of Chiefs of Police

Meghann Tracy: All right. Good afternoon, everyone. My name is Meghann Tracy. I'm with the International Association of Chiefs of Police in our Division of State and Provincial Police.

As Bill said, at our 2007 conference, we kind of began all of this work realizing that license plate readers were here, and we actually had an opportunity to be on the front side of the technology for once. We had seen what happened with red light cameras, with CAD and RMS. We didn't want that to happen with LPR.

Just a quick note, because my presentation does contain pictures: It is not a recommendation, nor is it an endorsement by the IACP nor the United States government.

OK. So why are we doing the work that we're doing? Because license plate readers are new, there's a number of policy and operational questions. There's questions regarding privacy. Some of the early adopters received a number of requests both from the public in general and the ACLU, and they came to us saying, “How do we respond, what do we tell them and is there anything that we really need to be thinking about?” And then, of course, there are the technical questions.

Just a quick overview for LPR, for anyone who's not familiar with it: The way it works is that the cameras capture images of a vehicle as it moves into the field of view of the camera. Different vendors do this slightly differently, but, essentially, the images are then converted into an electronically readable format using a series of rather sophisticated algorithms, and then it's compared against the agency's hot list. And this hot list is nothing new. It's something that agencies have always used.

And then just some of the information that's gathered by the license plate readers: There is the time and date. Most systems now have a GPS system on them. There's also images of the vehicle. There's a black and white infrared picture as well as a color photo, and depending on the system and depending on the agency's needs and requirements, that can capture anything from just a little bit of car right around the plate all the way to the entire image of the vehicle.

For our purposes, we classified them into three categories. You have mobile, fixed and then the portable covert. And the reason for this was as we got into our work, we needed to see if there were any differences in the way they were treated in the retention schedules and the way they were used, and you can see examples of each up there.

So one of the first areas that we looked at is the policy and operational guidance for law enforcement, and this is a project that's funded by the National Institute of Justice. In addition to this project, the other way we're addressing the policy side is through a model policy, which is currently under draft, and that's supported by the IACP's Law Enforcement Information Management Section in cooperation with our policy center. Once that goes through its final review process, that will be available through our policy center.

So the Policy and Operational Guide project: We wanted to look at how LPR was being used, who's using it, what were some of the lessons learned, the success stories. Essentially, we wanted to let agencies interested in LPR know how other agencies are using it and what are some of the things that they've already run into and dealt with.

A really quick background about the project: We collected our information two ways. The first was a certified random sampling of 500 agencies for size and agency type. We took a two-phase approach, and the first phase, we initially contacted all the agencies to introduce ourselves, to tell them about the project and to find out if they were using license plate readers. If so, we wanted to know who the point of contact was, so that we could send the survey to the best individual.

And then the second phase was actually delivering the survey to those who had responded yes, those who weren't sure and those who we weren't able to contact during the first phase.

So, initially, we got a response rate of … or, initially, we got 77 agencies saying, “Yes, we're using license plate readers.” However, of those 77, only 33 were actually able to complete the survey because they, for the most part, had just started using license plate readers. They weren't far enough along to really be able to answer the questions that we were asking.

We also conducted site visits, and we went to 12 agencies, and we had a couple requirements. One was that they had to have a 12 month implementation, and the reason for that was because we wanted to go to an agency that had a somewhat mature program who had faced some of these things and who could instruct us.

And just to give you an idea, these are all the agencies between the site visits and the survey, those 77 who responded yes that are using license plate readers. You can kind of see where the clusters are, and the reason you see New York that had so many, because their Department of Criminal Justice Services is really kind of leading the way and coordinating a lot of the state effort in that case.

So what were the main reasons for using LPR? Why are agencies getting into it? Stolen vehicles is the primary reason most agencies get into it, but that's not really where the true value of license plate readers are. Their true value is in the investigator purposes. Plates can be queried by plate, the full plate, a partial plate, location and time, address within a certain radius, if it's a fixed system or even if it's a mobile system.

And, as you'll see later, it really is the investigative purpose that's really the key thing here, and the with the third most common one was traffic initiatives, and there's a couple things with this. One, when you have a traffic stop, say you set up a DUI track point, it would be kind of nice to know if the vehicle that's approaching that checkpoint is a vehicle of interest for any reason.

During one of our site visits, the agency relayed this story. The reason they began using their portable system for their checkpoints was because two years prior, during the traffic stop, it was a DUI checkpoint, seat belt checkpoint. A gang member driving into this traffic stop didn't realize it until he was well within the line of cars. There was no getting out. Instead of just going through the checkpoint, he got out and started shooting. So now they want to know if someone is coming up. If nothing else, they can begin to prepare themselves.

And the other reason is … you see the quote there from Trooper Jonathan Olsen, “The quantity of stops has gone down, but that is because the quality of stops has gone up.” This is one of my favorite quotes from all the people that we talked to because it really shows the value. He's not pulling over random person after random person. They have a very targeted program in their state.

The people that they're getting are the people that they need to be getting. They have a rather large problem with unregistered and suspended and revoked drivers, and it's caused a significant … they're the ones causing a significant amount of the accidents in their state. So they're really targeting these people. They want to get them off the road.

Some of the big lessons learned from this project: Implement and train to policy. You need to get IT involved in the beginning. It seems so obvious to all of us sitting here, but it's really one of those things where many agencies went in with the idea that this is plug and play, and it's really not. There's a lot of things that need to be considered.

The use of wild cards. Remember to check your victims. When you're setting up a mobile or a portable unit, really target your hot spots. And there's an excellent study being done by George Mason University, which is also funded by NIJ, that's looking at this.

And then, if you're setting up a fixed unit, you need to think strategically. Just because you can put a unit up someplace doesn't make it the best location. If at all possible, move to wireless update and download. Your information is only as good as the last time you updated your system.

And when you get an LPR hit, there are two things that need to happen. The first is that you need to verify that the plate image matches the plate run by the system. The systems are getting better every day; however, there is still some difficulty in determining states, and also, if you have a system, if you have a plate that's ABC1234, that 1 is going to be run as a 1, as an L and an I. And you need to verify the plate status through your dispatcher MDT.

And one of the cases that really kind of illustrates all these points is one that came out of Jefferson parish with their sheriff's office. They had an ATM business owner and servicer who, during the course of putting the money in the ATMs and doing all those things, was robbed and murdered. They had a partial description. They had a description of the vehicle; it was a green Mitsubishi with gold rocker panels. So they went to their LPR system, searched, couldn't find anything matching that description, and they didn't have a plate number.

So, as they were kind of going through and, you know, running some ideas past each other, one of the investigators had an idea: Let's run the victim's plate. That was kind of their “a-ha” moment. By running the victim's plate and running a certain number of cars and times afterwards, soon they began to notice in some of these pictures, the victim's vehicle and then a couple after, there's this green Mitsubishi with gold rocker panels.

So, from this, they were not only able to get a plate number, they were able to show that this wasn't random; it wasn't chance. They were following this gentleman, and it actually did end up leading to a conviction.

Some of the emerging issues: We're seeing a trend of the static LPR cameras merging with in-car video. It's new. We don't really know what the accuracy and the effectiveness is, and we also need to know how this will affect the standards effort. There is a proliferation of private LPR scout cars. LPR has been around for a while on the private side and on the traffic enforcement. Really, it's relatively new to law enforcement, and there is some concern of what happens when you start mixing that data. And there are some data-sharing things, which Dale will talk about in his presentation.

There was also a number of privacy concerns raised, you know, the data collection, the access, retention, is this personally identifiable information, “the police are going to be tracking me,” “everyone is going to know where I'm going.” Well, plates exist to identify a vehicle, and because of this, because of all the questions that are already being raised, we formed a working group to develop the privacy impact assessment for the utilization of license plate readers. If you're interested in this document, it is available on the IACP's website. I'm also happy to send you a hard copy if you'd like.

We released this last year at our annual conference, and it is our finding that LPR data is not personally identifiable information. There is a step that must be taken in order to connect the license plate with the registered owner's information. It doesn't happen automatically. The LPR system does not query the DMV. It does not query your health records. It does not query anything else other than the list of plates that it is put into the system.

It may seem kind of funny that I say health records. Well, up in Maine, there was earlier this year some back and forth where the legislature was looking to ban license plate readers, and in the conversations that they had had with the legislators, they actually did think that it was going to query all of these databases. That's simply not the case. In order for that connection to be made, you must still do the investigative work.

At the time that we developed the PR report, there were no court cases. Recently, a decision in U.S. v. Wilcox came down that addresses this issue. The defense claimed that the LPR invaded the defendant's privacy interests. Wilcox was driving down the road, committed a vehicle infraction. His plate was run by the two officers driving. It came back as not belonging to that vehicle, so they pulled him over. Upon a driver's license check, they found that Wilcox had an outstanding warrant against him, arrested him. During … as they were searching through his vehicle and searching through his person, they found a couple of firearms on him, and he then had another charge added. Obviously, felons aren't supposed to have firearms; at least in Georgia they're not.

And so he said, “No. You need to throw this all out. The LPR violated my privacy. If the cop had looked with his eyes, it would have been different, but because he was using technology, it's a violation.”

The court said, “No. There is no Fourth Amendment protection. The law requires a plate display. In fact, it is the state that gives you the right to drive on a road. It is the state that requires a plate be displayed, and be displayed at night with that little light that illuminates it for the purposes of identifying that vehicle. And this has been around long before license plate readers came into being,” and they cited the rulings from the 5th, 6th, 8th, 9th and 10th Circuits. And, really, they noted that LPR is not a technology that invades privacy like thermal imaging, which is what the defense was claiming.

A couple of additional resources. I mentioned the standards effort earlier. The IACP has received funding from NIJ to help them in the standards development process. This is actually a three-phase project. Currently under way are the in-car video standards. Following that will be the interview room recorders. The third and final phase will be license plate readers.

There's also regional training that's being provided by the IACP to the TTAP project, which is a great training on the policy side, and then there is, as I mentioned, the George Mason study and the work that my co-presenters are doing.

There is a portal now available on the George Mason website. I am sure Bill can tell you more about that.

Please feel free to contact me if you have any questions, if you have any stories to share. We're doing a lot of work on the policy side, and I'm happy to talk more at length about it. There's a lot going on.

So now I'd turn it over to Dale.