As the second-largest metropolitan area in the United States, Los Angeles is home to four million residents and the mecca for most of the entertainment industry's high-profile events. The almost constant stream of celebrity-infused happenings that require serious crowd control keeps the Los Angeles Police Department very busy.
Knowing the city could always be target, the Counter Terrorism and Special Operations Bureau of LAPD several years ago began to seek out a surveillance system that includes cameras easily installed on a temporarily basis for covert investigations— as well as for pre-planned public events in order to manage crowd safety.
The department migrated from a cumbersome system that required officers to be nearby and did not allow for surveillance from remote locations, to its current system, which includes high-definition cameras that stream video to a remote command post and is also agile enough for use in densely-packed, large-crowd events. Rich Cowgill of the LAPD's Technical Support Unit details how the organization found the system it is working with now, and the challenges they met along the way.
CSO: What was your surveillance system like before you moved to the one you are using currently? In 2006-2007, I worked the narcotics division and we were using cameras with conventional microwave; analog cameras linked to a microwave transmitter. We were trying to hide them any way we could. We had to watch them remotely out in the field. An investigator would have to sit somewhere in the range of that conventional microwave broadcast, so a half mile or less away. The investigator would be sitting in his car all day with a little monitor about the size of a laptop on his screen with cables and it was a big hassle.
The signal was good if you were in range because it was like watching television, but you were very limited by the range. If you wanted to get some pans and tilts or other movement out of it, you had to use what they call a DTMF decoder, which is like what is on your telephone. You basically have a little radio with a keypad on it that would make phone-tone sounds and that would move the camera and it was all based on how good your proximity to the camera was. You had to be out there. We were looking for some flexibility to get away from that and network cameras were getting stronger and stronger. So we said "let's try a few of these." We bought 3 or 4 of the first PTZ (pan/tilt/zoom) network cameras that had come out and used those first.
What was your initial impression?
Right around that time we also started implementing our first internet-based cameras. We were going through a microwave, point-to-point network. In other words, we would put a camera up somewhere, but along with the camera we had to put a transmitter and literally a satellite dish; something similar to what you see for satellite television. We would aim it at a mountain top where one of our network providers had a backhaul into the web. We would hopefully get the signal. That was real time consuming and not very effective.
With the new network cameras we have invested in, we put them in an enclosure and made it into to one of those satellite dishes. We had a strong link, we had excellent network connectivity. For the first few days it was working really well. But then the camera went off line. We had to go back out and reset it. It was just a simple reset, but we didnt have any remote way of resetting the camera at that time. We got it back online and it was working fine, but then it happened again. The case the camera was being used on was a complex narcotics investigation and we had to get a lot of intelligence. The camera was going to be in place for a long period of time, we were thinking a couple of months, maybe longer. And within that couple of months, we had to go out to that site and reset it probably about a half of a dozen times. It actually compromised the investigation because they figured out what we were doing.
We tried to get some tech support from the vendor that manufactured the camera and they were pretty unresponsive as they had been the whole time we were going through this.
So, we started looking in another direction and stumbled upon the Axis cameras from one of our distributors. He recommended Axis — the 232 was the main camera at that time and we bought 3 or 4 and built them into some enclosures. We found we had no problem right away. The connection was solid and strong. We didn't have one issue with it throughout the entire investigation so we were sold at that point. The cameras stream encrypted video to a camera station video management system (VMS) running on a server at the command post.
Have you experienced other challenges and made modifications along the way to your surveillance?
One change we made was we migrated to a more dependable and secure mesh network. We've become more involved recently in situational awareness video, and providing video for major events like the Academy Awards, the Grammys, the Lakers, who have won the NBA championship a few times. The team has had large parades and we would put cameras up along the route.
For events, we were using cameras based on a cellular network on the back end and a router and maybe a cellular air card built into it. But we found that any time the cameras would be online during these large events, because so many people use cell phones now, when large pockets of people would come through the area, cell phone use would interrupt the video stream. The camera would slow down and go down to a frame rate that was just unacceptable. We might see an image on the screen but you would look at that image for a full minute before it would go to the next image.
We knew it was a problem, but it wasn't until a Lakers parade in 2010 when we knew we wanted to go with a dedicated separate network, like a mesh system. We had some resistance to that idea, but they still wanted us to put up our cameras. We knew they were going to fail [without a more robust network]. We put them up anyway and we had the same issues. There were a lot of questions in the command post that day about why that was happening. That is when we turned the corner and got the authorization to go with the mesh system we currently have in place, running high-definition cameras on mesh nodes.
What are you doing with the information the cameras provide for you?
When we are doing a high profile event we are looking at crowd size, crowd behavior, trends, what they are doing. We had the Prince and Princess of England out here last summer, and we had the mesh nodes deployed and Axis cameras around the venue they were attending in downtown LA for a few hours. There were huge numbers of people out; they had barricades in place, but there were still so many people moving throughout the area.
It was very helpful, and I think comforting, to the commanders at the observation post to be able to get a feel for how big the crowd was and what they were doing. The cameras were able to move quickly, they responded quickly, the video was clear, the frame rate was great and bandwidth was not a problem.
It's kind of security blanket for them when they run these operations. The same thing happened during the Occupy LA movement. We had a mesh system with Axis cameras deployed around the encampment at City Hall Park late last year for most of the last few weeks it was going on. When it came time for the department to make its move and clear the park out, we had total situational awareness of what was going on in the park; if there had been any movement of people with weapons (which there were not), it was a way to keep commanders up-to-date and give them that bird's-eye view.
When we went in and removed the protestors, the video was able to record that the officers doing this were conducting themselves professionally and that there weren't any issues. If there had been any allegations of abuse, we would have had the video.