Are You Throwing Money Away?

It is not a video surveillance system—it is a data collection system

• By Scott Seraboff
• Feb 19, 2019

Around the world, there are literally millions of video surveillance cameras. Whether analog or digital, these cameras are recording video (and sometimes audio and metadata) that covers more information that can possibly be seen or understood at the time of the recording.

We then see these systems delete this information, on average, after about 30 days of collection, and start over again.

Even within the small sample of one 30 day period within one closed system (i.e. a large drug store) there is information within that captured video that is valuable. This is not “forensic value,” meaning information that allows us to go back and discover something criminal or illegal and then correct and prosecute, but rather value inherent in business intelligence. This can be information on customer patterns, how long the customer lingers in one area, or an untold number of additional bits of information that can make a business smarter and create revenue.

We, as users of video surveillance systems and installers of these systems, are missing a critical piece of the pie with this video data. Yes, we are seeing the cameras within these systems collect video data, but to think of them as only cameras is to miss a big part of the picture. They are not cameras, they are Data Collection Devices, and with each passing second they collect data that can be used by the business owner and by third parties who wish to take that data and analyze it to the benefit of the user and potential third parties.

Thirty cameras over a 24 hour period, over a 30 day period, will record 21,600 hours of video. While this is a statistically small number, imagine if you could expand this collection to a year in the one location. Now you have 259,200 hours of video stored, and this is statistically much more significant than one 30-day stretch. Within that 259,000 hours are tidbits of information that can and will be valuable to a wide range of potential users. It is not hard to imagine the uses of this data for the business owner; what is harder is to try and wrap your mind around the uses for people outside of the end user’s business. If this was a single location of a major drugstore in Dallas, the collected data could and will be valuable to every single company that sells product through that store (an example of this is data collected at a Wal-Mart; while Wal-Mart had no inclination to use the data, a drug company that sold through Wal-Mart did and was able to increase sales through data analysis and then a change to the way product was placed). If our hypothetical drugstore had cameras facing the busy intersection they might very well find themselves in the same position as a small, family owned trucking company who was paid $52 million (Australian dollars) for access to traffic pattern records they had kept since the inception of the business.

This is a real potential source of revenue that is being thrown away by each and every user of a video surveillance system. The collected data is theirs, and as such they are throwing away the potential to turn that data into real dollars and cents when they allow this data to be deleted after 30 days. When was the last time anyone thought of a VMS and security system as a revenue generator? 

A VMS, and the systems that go with it, is an expensive investment. The storage component of this expense is determined by the amount of data that will be stored. The longer you want to store data (and this will take into account quality and framerate), the more the storage expense will grow. This is the reason that systems are still thought of as a “VMS” and not a data collection system (DCS) with video, audio and metadata as the collected end product. As much as our local supermarket would love to store data for 3 years to match the state slip and fall statute, the cost is simply prohibitive. To add to that problem, finding space for the hardware and then maintaining the system simply makes it harder to reach that lofty goal and so the majority do not even try. Even the use of the “cloud” as the system is expensive and also adds the cost of bandwidth needed to deliver video streams to the cloud.

Think of the system as a DCS, and realize that it can deliver value to the end user and even real revenue, and you stop thinking of your current surveillance system as just an expense. You can empower each camera on your system to become a data collection device with potential to deliver massive returns. All of this can be done while simplifying the installation of the system, decreasing the amount of equipment needed, and dramatically reducing the need to spend capital expense on hardware. Instead, a minimum capital expense can be undertaken, and instead the entire system can be run as an operational expense with a monthly cost. Data collected can be stored forever and the end user can participate in revenues derived from the rental of the data to approved third parties. Even if a system is currently in place and local storage in operation, all of the above can still be done with nothing more than a simple add-on to the system.

This is your data, plain and simple, and it is an asset that is yours to use. You own it. You ought to see a return from this that is no different that having a retail consultant suggest new items for the “impulse buy” rack and seeing those products sell faster and at higher revenues. You are allowing money to get erased after the end of a retention period, and this can be stopped with little expense and without upsetting a current system.

Cloud storage is clearly a part of this positive solution, and yet has a reputation for being expensive to deliver data to the cloud and return from the cloud. Of course, many cloud services want to be the entire front end- they want to take your camera at 30 fps and 2MP and store it for you. With this, the cost to store seven days is somewhat high and the idea of providing 30 days, to say nothing of three years, is out of reach. This does not show any creativity. Why store at 30 FPS and 2MP for 30 days when you can reduce both of these dramatically to save on storage costs within the cloud? Why not focus on high res for seven days and then have a dramatic decrease after those seven days?

As an industry we have been taught, and almost wholly brainwashed, into thinking that storage has to be expensive. We have been told that the cloud doesn’t work for long term IP cameras, or large quantities of IP cameras. We have been trained to see that we collect video from the system and, if nothing bad has happened, to toss that video (this is a lot like the idea of motion detection, which was created to save on storage) and start over. This is old school thinking and while it has served us well, new technologies and new systems have evolved to allow us to store data for as long as we wish, and to use that data to create real value for the end user and partners.