Digital Video Surveillance Best Practices

Tell a Friend

Michael Dillon, Vice President of Business Development for Firetide, Inc., chaired the roundtable Broadband Wireless, Public Safety and Video Surveillance at the 15th Digital Cities Convention in Washington, DC, December 11–12, 2007.

To kick the session off, Ray Cooke of IBM Global Technology Services reviewed the realities and requirements of video surveillance and what people can expect. What is the tradeoff, for example, between image resolution and quality within the video and the amount of storage or bandwidth required to deliver it? Visit the Proceedings page to download his and other presentations from this session (log-in required).

Three Video-Surveillance Case Studies

Case Study 1: Jubilee Park and Community Center Corp. is a nonprofit organization driving community redevelopment activities at Jubilee Park, a high-crime community in South Dallas. Firetide wireless mesh and Sony cameras have been installed at intersections throughout the community and are controlled by the Dallas Police Department, providing police with the capability to monitor activity in real time.

"Jubilee Park is an excellent example of what I like to call Muni 2.0," Dillon said. "This is a community group reaching out and extending and expending real dollars to build real solutions on a real time table that gives real results, and from that you get the organic growth effect of saying there can be other projects that show that this public-private collaboration can work."

The deployment sits considerably outside a separate network of cameras in downtown Dallas and sets up a model for how this might be taken out to other communities and grow in Dallas.

Case Study 2: Detective Chris Jensen talked about how the Phoenix Police Department was actually able to capture a serial killer using covert digital video surveillance. Detectives and technicians suited up to look like utility workers to install the infrastructure. "They will work an area and in some cases drive the bad guys to areas they think are secret and private where they can have their communications and record that," Dillon reported. The system is also portable and may be moved around.

The group reviewed different kinds of surveillaince across three categories:

1. Overt uses big camera pods clearly identified with police badges or some other authority or presence. It says, "Hey, we’re here, we see you."
2. Covert is used when you don’t want perpetrators to know the camera is there. "You want them to behave openly and naturally, and you want them to pick it all up and take it back to the DA and prosecute,” Dillon.
3. Transactional is used tactically; for example, at a department store, where cameras are sitting in the ceiling. Security personnel may or may not be monitoring them or have any intelligence on the back end, and the footage may or may not be stored. "But situationally and tactically, when somebody needs to go see that video, they walk over to it and view it," Dillon said.

Leyva referred to "frequent fliers" — people who claim to be sick and are looking for a bed to sleep in for the night. Tucson has been able to cut down on this use of the hospital resources because the doctors and nurses can look on camera and see that it’s a frequent flier. An assessment is made that there’s nothing wrong with the patient, and he is removed from the ambulance.

In other cases where there is trauma, doctors and nurses are able to deliver a level of care through the hands of the paramedics while the ambulance is actually under way. Seconds save lives wth cardiac arrest and situations where there might be arterial bleeding. If you can look into the vehicle and can tell medics what to do by authorizing treatment early is definitely a life-saving technology and another use of video.

Dillon said: "The same type of infrastructure that is used for digital video suveillance can be used for traffic controllers; hence the reason that traffic engineering is the primary player and host of this system. Public works and traffic are making the poles their signals are mounted on intelligent so that you can use that for a public safety and medical applications like health."

Digital Video Surveillance Best Practices

The case-study presentations sparked considerable discussion on best practices, multipurpose networks, and the need for federal policy direction on implementation.

Approach: There was debate about how to approach projects: Do we spend rigorous amounts of time trying to plan for video, or do we basically find a project we want to do and jump out there and do it?

"At the end of the day, you need to understand the user and the use of the video before you do anything, and that should be the primary driver to design. In essence, form still follows function."

If high-resolution video is required to see license plates, because that’s the ultimate goal, short cut on the bandwidth, image size, capture capabilities, or camera type can't be taken. "It may negate the video if I can’t use it for the intended purpose."

A learning curve seems to accompany most of these projects, with project leaders evolving into the correct solution as opposed to striking out and getting the right answer and implementing successfully right off the bat.

Secondary and tertiary benefits can often emerge as the real drivers behind a deployment. Dillon said: "Yeah, we put a camera up, but the camera’s up not just to see but to make the street safer, so that people will come out and revitalize downtown. It’s up so that when people come out, they can go to the ATM to get money to spend in the store and not get robbed or have to deal with drug dealers and prostitution and things that are going on because of the current effect of it."

Multipurpose Networks: In video, most of what you are doing is backhauling data versus the typical public-access or government use, which often involves sending up very little data and bringing back a lot of data. There is an inverse relationship between what video is doing and what typical browsing and other applications that are used on the Internet and other devices are doing, and they may have a symbiotic relationship and coexist comfortably.

"The possibility of mixed use, shared use, multiservice networks is alive and well," Dillon said. "It is not at all something that is impossible, it just requires the appropriate QoS, security and, for that matter, partititioning of that network to afford everyone the needs they have to afford the security, privacy and, for that matter, security of the data."

Socializing Video Surveillance: What are the best practices? Dillon recalled a time "when we were dealing with the Orwellian concept — Big Brother is watching . Every time you heard about a camera system going up, you thought, they’re invading our privacy."

There was a mixed review on who was experiencing a paranoia from the public at large and as well as from stakeholders on deploying video cameras. The best practice is to socialize video with the key stakeholders and the community beforehand. This is done to block the potential that the people who still feel that way will go to heroic efforts to try to put down the project.

"Instead, it’s better to make sure that you have them on board, allayed their fears, and in controlled settings that you’ve let them deal with their paranoia of video. Get that under control and then go forward with your project having already put the fire out."

Policy Needs and Opportunities: The federal government needs to own published guidelines for video surveillance — broad guidelines, drivers to say this is how you should do your policy. "We believe probably at the state or county level this would be where this would normally come to roost, but that federal guidance would promote continuity from jurisdiction to jurisdiction, geography to geography, on how policy is implemented from city to city, county to county, and state to state."

Video surveillance policies should be uniform as well to be sure you’ve already put out the fires before you can go down the path. If this type of policy is present and already vetted, then the implementation of the camera and the value associated with it will not be subjected to that jeopardy of having folks not wanting the video there.

Especially with telemedicine and in-car video for police and fire departments, there is understandable concern that with a camera there watching every move a practitioner makes, are they going to spend the rest of their career carrying a workload of lawsuits based upon a mistake or misdirection by a doctor? Good policy needs to include indemnifications and protections for both the users and even the providers. On a broad basis, if these protections can be provided, the video can mature to its full value.