In an article just posted today from Indy Week, an independent news and lifestyle publication serving Raleigh, Durham and Chapel Hill, N.C., the reporter gets into the nuts-and-bolts of why a city's municipal surveillance system didn't quite work as planned. Read the article, "Nobody's Watching: Durham's police surveillance camera system barely functions."
Based on what we know from this article it looks like these were the weaknesses:
1. The data network available for the video surveillance system didn't have the bandwidth that it was said to have had.
2. The network for the surveillance system was shared with other city functions.
3. The city didn't commit the kind of budget that it needed to establish and maintain such a project. (The original budget was less than $90,000 -- a comical number for a 13-camera system that is supposed to push video wirelessly to police cruisers and the police command center). The only originalÂ bid was over that amount, and then theÂ integrator was later forced to bring their cost down to under that budget.
4. The wireless network system was plagued with lag (delay).
5. The integrator had to eat plenty of cost-overruns.
I'm not sure there's any blame to be had here, although the IndyWeek.com articleÂ seems to want to dredge up the blame game, and the city seems intent on blaming the system integrator for the city's own lack of bandwidth. What you have is an integrator trying to do a project on a city's shoestring budget, a city that belived it could buy the moon on a tiny budget, and political dreams that a city surveillance system is going to be a panacea to stop crime.
From covering this trend here at SecurityInfoWatch.com, there seem to be some real strategies that work well. Here are some tips I've picked up along the way from talking to vendors, integrators, wireless companies, camera companies, cities and even police officers.
1. Figure out what bandwidth you need and then increase that some more. Video eats bandwidth. Repeat: Video eats bandwidth. Some vendors will tell you bandwidth is not a problem, but that either means that 1) you have a big fat bandwidth pipe, 2) they don't understand the bandwidth needs of the system they are proposing, 3) they don't know much about your network and its available bandwidth, or 4) they've actually done their homework and areÂ planning on adding bandwidth to accommodate the system.
2. Go in with a comfortable budget and be realistic about what such a system costs.
3. Plan for a testing time to work out the kinks.
4. Expect lag on a wireless network until you can provide lots of free bandwidth and a very few number of jumps.
5. Dedicate the networkto the surveillance system. Horror stories happen when you take an existing network, with its own demands, and expect to put a ton of video on top of it.
6. Scale up. Plenty of cities start with only 3 or 4 cameras and then scale up as their budget allows, but they design the system -- even when it only has 3 or 4 cameras -- for that day when they may have hundreds.
7. Don't get stuck with your camera placements.Cameras pointing at the east side of a street means the drug dealers then move to the west side of the street to move their dope. Smart cities have designed their systems with wider angle lenses, PTZs, or even easy-to-move installations. That's the great thing about wireless -- you just move the camera to your new location and establish power. You won't need to drag the coax or the Cat-5/5e/6 cable.
8. Budget for maintenance and equipment replacement.Â Â Outdoor environments are hard on electronics, no matter how well enclosed they are. Condensation, heat, cold, etc., all eventually lay waste to materials. Concrete bridges eventually crumble, and surveillance systems aren't made from a material even as durable as concrete.
9. Don't be afraid to walk away from a project. If the end-user is not realistic, don't climb into the pit with them. In this case of Durham, it sounds like the mayor was grandstanding to make himself look like a "tough on crime" type leader. This project sounds like it was doomed from the start.