Joel, are you most interested in storing the body camera videos or?
John, I am thinking that ideally the storage of substantially all of the recorded video from public devices ( body cameras, street cameras, municipal facilities, etc.) will be federated as a single storage array with massive capacity. Otherwise issues of unified access and use of recorded video as well as maintenance will be suboptimal to point of not being acceptable.
Hi, we have our VMS which we did use so far only for our own projects. Planning to send it soon to IPVM for testing.
However, we are pretty good in large scale systems with up to more than 10.000 cameras. We are working with DELL and I would recommend it. We have integrated it with Dell and should be certified as a Software solution. I do not prefer one large storage sytem, it should be modular. With the integration with Dell all Servers and storages are treated as one Pool. The user doesnt need to take care of the distribution of the cameras on the server and storages. Its all done by the system, including automatic failover and similar.
Joel, about body cameras I think its an serious issue to access the records and store them in one centralized data base, accessable at anytime from any authorized person. Hand carrying it by SD card I believe is not usable really. Your body worn camera should have wifi streaming or any other automatically usable uploading method to the system in the car or any other place. We have develope an integrated solution which can record the data on the car and than anytime when connection is available synchronize it with the central database. You'll be able to find the videos then later by the ID or the gelocation.
Is the Secure Cities initiative to which you refer a particular, well-defined initiative or are you talking about general trends around municipal deployments?
In either case, you might be jumping to some conclusions about specific requirements that need be met, when in fact they may not be requirements. You mention retention for 2 years of 10-50,000 cameras. Where do these numbers come from? Is this for all video from all cameras or for video ‘tagged’ or ‘marked’ as requiring retention? When you talk about body camera systems, are you sure they need to be deeply integrated (using the same storage subsystems) into the systems recording fixed cameras or can they be integrated ‘on the glass’?
Divide and conquer is often a good strategy when dealing with projects this big. Is it possible that the large municipal systems you’re talking about can start with high risk and high value parts of the city and then conglomerate or federate both administratively and technically?
I do agree that large camera count deployments challenge storage. But the further in the future the deployments are the more Moore’s Law helps things along. Furthermore, complex systems like this will likely provide some give and take in terms of what’s practical (both in terms of cost and technology) and what’s required.
Btw, excellent article about storage costs of police cameras. Key quotes:
"That's the biggest problem with this system...the cost of the storage," Brewer said. "They do offer unlimited storage, but it's quite costly -- well above $1 million for the package we had looked at."
"But it's not the cameras that generate the most money. Glenn Mattson, who follows Taser as an equity analyst for Ladenburg Thalmann, said the company makes a far bigger profit on its storage service than hardware. Last year, Taser's gross profit margins on hardware were 15.6%; the gross margins for video storage were 51%, Mattson said."
"There's no contest. They don't care about making money on the cameras," Mattson said. "If they can just break even on them, it's fine, because they're going to create this high margin stream of revenue on the video side."