Wireless Government

Subscribe to Newsletter
Tell a Friend
Print this Page

02/28/2008

Mesh-Enabled Video Surveillance and the Phoenix Police Department: Desert Guardian, Super Bowl XLII Case Studies

Detective Chris Jensen with the City of Phoenix Drug Enforcement Bureau has more than 23 years of law-enforcement experience. He is the systems administrator for all electronic surveillance platforms for the Phoenix Police Department, including both audio and video surveillance. Chris helped design and install Phoenix’s video-surveillance system, Desert Guardian, and oversaw downtown surveillance operations during Super Bowl XLII this past January. Det. Jensen participated in the W2i Webinar “Public Safety and Mesh-Enabled Video Surveillance," sponsored by Firetide, on February 27, 2008. The following is a transcript of his remarks.

The history of video surveillance in Phoenix has been a function of our covert investigative details, including the Drug Enforcement Bureau. Our old system was analog-based microwave, and what we experienced was pretty terrigble image quality, especially if you were more than a few blocks — or sometimes less than a football field — away from our cameras. It was seldom used by our investigators, especially in the summer, because it gets really warm. It wasn’t really user friendly to be out in your vehicle when it’s 114 or 115 degrees, trying to monitor a camera in a car. We could never use this type of video for patrol operations. Maybe once in a great while we could find a use for it. But it wasn’t really high on our priority list.

As far back as 2004, we spoke about going to an IP-based video system. The biggest resistance we had at that time was from our technical staff. I think that’s a universal in the law enforcdement world — that we really don’t like to change unless we’re either forced to, or we find something that works a lot better. In late 2005 and early 2006, we had a citizen's narcotics complaint that we were looking into. The citizen got a little tired of us coming out and not being really responsive, so he actually set up a webcam of his own that looked in the alley way he was complaining about. We used that camera and successfully solved that problem for that neighborhood. We then began discussions about moving to an IP-based mesh system. What we found in talking with our city IT and departmental IT folks is no one really had any experience or knowledge in either of those technologies.

Desert Guardian

In 2006, we really were in the high swing of investigating two serial-killer sprees. One was our "Baseline Rapist," and the other was our random shooters, who, between the two, almost on a weekly basis were either injuring or killing somebody in the Valley here. We were mandated by our chiefs in our city hall to get some video up and running to assist in this investigation. We were in the midst of just going through determining what technologies we wanted to use, our integrator and our suppliers, when this project kicked into high gear. We were able to establish about a 30-camera network over five mesh networks in about a five-square-mile area of Phoenix.... Out of the box, this technology does work. Obviously, there were times we had problems in that two-week period, but we were able to overcome that and get this network up and running quickly, including the infrastructure.

When our integrator was installing this, one of their engineers said if we could catch this guy on video, that would be wonderful. My comment to him was if that was our measure of success or failure, we’re setting ourselves up for failure. [It's important to] have some real expectations and specific goals in mind that you want to achieve with your video project. Because my goal was not to catch the killers with the cameras. My goal was to get the video cameras up and running to free up surveillance officers who would have to sit at those 30 locations. We were able to free up nearly 30 surveillance officers who could be used in other parts of the investigation to follow up on other leads. I think that’s what our success was; that we were able to arrest these suspects more quickly because we were able to free up those resources to follow up on those other leads.

That really was the start of our IP video surveillance system that we call Desert Guardian. The Phoenix model, since we’re driven out of an investigative bureau, and more of a covert investigative bureau, is event driven. We normally do temporary deployments, and to this day we don’t have any permanent installations. We normally are covert. Our deployments are geared to achieve specific goals. Before we hang a camera, we will review with the investigator or the officer who wants a camera set up and ask them what their goals are, how they’re going to use the camera — not only for intelligence but operationally. That way, before going in, we know we are going to be successful, how to use video correctly in and investigation with specific goals. Once those event goals are met, we redeploy our resources to meet the needs for other investigations or events. That way we continually use our resources that we first brought on board in the summer of ‘06, and we expand our resrouces as needed, but we reuse those resources over and over again to give the city more bagn for the buck.

Design Considerations

Obviously, we did things a little in reverse because of the emergency situation we were in. Our biggest design consideration was flexibility. We used our network video recorder, mesh infrastructure and IP cameras. What we found is that these three things gave us the greatest flexibility. It doesn’t matter what type of camera we are using, we know it’s going to work with our mesh infrastructure, and our network video recorder is going to bring it back and store that data for us. One of the biggest considerations also in any system is your backhaul design. We’re fortunate here in Phoenix that we have a pretty robust fiber network that runs throughout the city. We like to use that whenever we can, it’s free, we get the bandwidth we want, and it comes back. If we’re unable to use city fiber, we’ll use our local phone companies for DSL or our local cable company for cable Internet. As a last resort, or a "quick and dirty" camera, we’ll use an EVDO router and a broadband card to stick out with our camera just to get that signal back until we can get a more permanent solution. We have three different monitor rooms – kind of a baby bear, mama bear, papa bear set-up. We use the large monitor room for both audio and video surveillance, and there are 17 work stations in this room.

Super Bowl XLII

[The NFL] came to us last fall wanting some video surveillance. Our area of operation was the downtown Phoenix area, while the game itself was in the Glendale, Arizona. In the downtown area, that’s where the NFL headquarters was located, the media center for the world, and numerous other NFL events in that area — some pretty high-profile events that we had to protect.

We ended up deploying 42 cameras over five separate mesh networks. We established two operations centers — downtown and joint operation centers — and we were able to connect the two with about a little over a three-mile wireless shot. We used the Firetide mesh again for that, and we were able to link those two. Any video our downtown operations center had, that’s what we were able to send to the joint operations center.

We tried out a few new technologies for the Super Bowl. We had some rapidly deployable cameras, including some tripod-mounted cameras underneath a mesh node.... We used those to fill in gaps where our cameras couldn’t reach and for kind of quick and dirty deployments for events that popped up past the initial installations. Between the two operations centers, we established encrypted VoIP phones. Both the video and the audio at the same time rode over the mesh networks, and we didn’t see any loss of voice or video quality. That’s one of the things we were testing, to see if we had any degradation on either one of those.

In Phoenix, we have a lot of heat issues, so we’re trying out some camera enclosures that have some chilling technology in them to give protection past 140 degrees in the summer, which we need. These things also have heaters in them for those climates that need to worry about the other end of the spectrum.

Even though they talked to us in the fall, by the time we got our plan together and our purchase of the additional equipment we needed for the Super Bowl, we started our deployment in the second to third week in December. Take out the holidays, and we were up and running with the 42 cameras by the third week January. So we were doing our testing and getting our mobile command center set up in our downtown area. The other thing we were able to do is mesh-enable the tactical vehicle that our SWAT team uses, so that if they needed to come into the downtown area for a callout, they would be able to pull any of the video and also have the VoIP communication set up as needed.

What Makes Successful Deployments?

When we look at planning a successful deployment, the one thing we do is define our goals and expected outcomes. By going into a situation and telling someone you’re going to reduce crime 25–30 percent in an area, you’re setting yourself up for failure. I've read those in some case studies...when that doesn’t happen, people point at a project and call it a failure. What you want to do is specify you want to cover X amount of areas with X amount of manpower. You’re saving this manpower and you're able to redeploy that manpower to do other things. Crimes are of a random nature a lot of times and to catch them randomly on a video camera, I don't think we can actually predict that. But we can predict things we can actually control.

We don't only do a technical site survey, we do an operational site survey. We’ll go into an area, we can put cameras here, we know we'll drive whatever illegal activity is happening to another area. So if we're doing an overdeployment in one area, we may also throw up some coverts in a different area because operationally we know we're going to force our bad guys into a different location. So we look both technicdally and operationally when we do a site survey.

One of the biggest things you need to do is identify your backhaul infrastructure.... If you can’t get that picture back to where it needs to be used, you're not doing anybody any good.

Another key thing that we try to do is communicate all our tasks to all the parties that are going to be involved, including our integrator, IT folks, and our operational people within the P.D. In the last two years of establishing operational policy and training, I think that's key. You hate to have a great camera system or have some great evidence and have it thrown out because your operators didn’t know what they were doing or violated a court ruling or didn’t do things quite right. So from end to end we want to make sure...the more training you can do with your investigators and operators, the more successful your video is going to be.

When we look at our successes, one of the things we’ve worked hard on is our relationships. They’re essential. No one city department owns the entire infrastructure for a city, so a lot of it is relationship-building, partnership-building between departments that may not normally work together on a daily basis, whether it’s with our IT folks, our streets department, our water – I don’t care who owns the infrastructure as long as I can ride on it.

I really think the keys to our success are using proven technologies, using proven integrators and suppliers, and having the proper infrastructure. When we look at technologies, everybody still wants to do the proof-of-concept thing because this is all new technology to everyone. The farther down the road we go, I think we have enough evidence with the deployments that are out there that we can get a little sense of what’s going to work and what's not going to work, what’s becoming the industry standards in law enforcement. The proven integrators and suppliers — again, either word of mouth or looking at the other projects they've worked on — I think is a good indicator of success and failure. The infrastructure — I’ve spoken with cities, they have all the right pieces but they just can’t get it to work, and then you start talking to them, and you find out they really don’t have the right infrastructure to support what they are trying to do. I think those are the things you really want to look at going into any video project.

back

Post new comment:
Only registered users can add comments.
Please Log-in

• MORE BLOGS

Wireless Government

Mesh-Enabled Video Surveillance and the Phoenix Police Department: Desert Guardian, Super Bowl XLII Case Studies

Authors