A recent study of North American loss prevention professionals turned up some very interesting information on how CCTV is used in retail today. This independent survey, commissioned by Genetec, was conducted by the Loss Prevention Research Council (LPRC). Survey participants were executives ranging in title from Director to Sr. Vice President of Loss Prevention or Asset Protection. The type of retailers spanned the gamut to include Mass Merchandise, Department Stores, Home Centers, Grocers, drug stores and specialty retailers.

The first question asked was “are you using recorded video in your stores, distribution centers and corporate offices”?  It’s no surprise that 96% of the respondents use recorded video as a method to protect the people, property and assets of their companies.

When VCR technology first became available retail asset protection quickly leveraged the ability to record their surveillance cameras. Why? There was finally a method available to forensically investigate events that had already transpired.  This new tool brought challenges as well as opportunity. I can remember walking into store manager’s offices and seeing entire walls of VHS tapes. The least amount was 30 - one for each day of most months.  Now just because they had 30 cassettes didn’t mean that those tapes were rotated on a daily basis. It also didn’t ensure the tapes that were used contained quality video imagery.  The average life expectancy of a VHS tape was 8-12 recordings on a time lapse VCR – and that was if you didn’t fast forward or rewind very much.  I’m sure many of you reading this article can remember the adventure of watching a VHS recording that looked like it had been made in a blinding snowstorm, or that vertically flipped the images like a slot machine. Those same investigators wore out the tracking wheel on those VCRs trying to get usable evidence.  The worst part was when a tape became caught in the VCR head. As a cassette ejected from the machine the tape would start to unwind like a confetti streamer.  If you were lucky enough to get the VCR apart and the tape out without breaking it, you then had the task of rewinding it back into the cassette with any object you could find to turn the wheel.  Good times.

VCRs themselves had a fairly short lifespan. The recording heads were typically rated for 12 months. At the end of a year you had two choices – try and have the heads replaced or just buy a new VCR. Most loss prevention teams opted for the latter, although some system integrators later created “pools” of VCRs in loaner replacement program.

Needless to say, when DVR technology first arrived as a commercial alternative to VCRs in the late 1990’s it was embraced by retail loss prevention. The DVR is basically a video server with a built in hard drive. Eliminating the daily need to replace tapes was the single biggest benefit DVRs offered. It also eliminated, in most cases, the need for switcher, quad or multiplexer to encode multiple cameras.  The DVR also came with proprietary software for programming it to record at specific times or other triggered events. Installation of the device was fairly easy; cameras were connected via coax cable with a BNC to the back of the box. It was plugged in, turned on and in most cases ready to record with a minimal amount of set up.

DVR boxes were sold in a variety of sizes and configurations.  The most common was the 16 channel version. Sixteen became a prime number for retail loss prevention when designing surveillance systems for their stores, mainly because budget restrictions often precluded the purchase of additional DVRs.  I’ve seen some ingenious ways retailers tried to work around this restriction.  Four camera quads and switchers were sometimes looped into that 16th port to increase camera counts at minimal cost. The resulting video may not have been the best, but it was better than no camera coverage at all. Another restriction that retail worked around was the hard drive size. Since the box was preconfigured with 250, 500 or 750 GB of storage cameras were programmed to record on schedules, or motion to save drive space. The limits on storage also affected the quality of the video. Real time video is defined as 30 frames per second, yet many DVRs today are recoding at 6, 3 or 1 frame per second (fps) in order to meet the30 day minimum retention requirement of most retailers. More degradation to the video occurs in the resolution or downsizing of the stored images. Finally proprietary compression applied to the video dropped quality even further.  Still, compared to the VCR, the images were better, daily interaction was minimal and reliability increased significantly.

DVRs also offered another capability that Loss Prevention had previously not had in their arsenal – the ability to connect remotely to the store. The late 90’s were a renaissance for networks and the internet. Personal computers, ISPs and modems allowed immediate communications and information flow. Loss prevention needed to be able to see video quickly and DVR’s allowed some access to critical events along the information superhighway.  Well, sort of. Early video compression technology wasn’t designed with network access in mind. When plugged into store networks, the amount of data streaming through the switches sometimes took down the entire store network, including the POS. This didn’t endear DVRs to the IT or Store Operations groups. 
DVR’s in some cases are still prohibited from residing on the retailer’s network today. In the LPRC findings, 84% of those surveyed have remote connectivity to their DVRs. Of that group 60% are operating on their company network. But surprisingly, 24% of retailers reported using consumer grade DSL or Cable to connect to their DVRs.  Today 96% of those surveyed currently still use DVRs in their stores.  Almost all reported using multiple brands and versions throughout their enterprise. Remote connectivity is important enough that 76% of those surveyed said it is an absolute requirement to achieve their team’s objectives.

The LPRC survey revealed some other interesting information.  Multiple questions were asked rating the level of importance for various aspects of recorded video. The highest response to any question addressed the issue of reliability. 96% of loss prevention professionals rated system reliability as a key criteria in their selection of a recording platform. This response wasn’t unexpected, since what good are cameras if they aren’t being recorded? As reliability relates to the DVR, most have an average lifespan of 36 months. Many of the LP executives I’ve worked with confirm three years as average. Far more problematic, and at the root of the reliability response, is that when a DVR goes down, all recorded video is lost.  Murphy’s Law seems to come into play whenever one dies – it’s almost always at the worst possible time. Many times I’ve heard Asset Protection lament the loss of crucial evidence during an investigation. Unfortunately, there does not seem to be services readily available to retail that can retrieve those hard drive images.

Another question was asked that pertains purely to the recorded images themselves. The LPRC survey showed 84% of respondents rating the quality of video as an absolute factor in their CCTV platform. Over the past decade we’ve seen the analog camera manufacturers reduce chip sizes in order to minimize the camera size, reduce costs and minimize their footprint in the store. From a technology standpoint 540 total vertical lines (TVL) is the highest resolution available in analog cameras today. Analog cameras also rely on interlaced scanning which tends to blur moving images.  This is most notably seen when capturing a still image from video with movement; like a car driving away or a suspect running. Add the limitations placed on the DVR in terms of frames recorded per second, resolution and compression and it’s understandable why retailers put such emphasis on video quality.
Now that retail has identified some of the most important needs in their video capabilities; image quality, remote connectivity and reliability what technology exists that can fulfill these needs and provide additional benefits? Before I answer, I want to add one more criteria – Price. 68% of those surveyed indicted price was a major determining factor in their technology choice.  So what technology can satisfy the requirements of retail and still not break the bank?

It’s IP video.
OK, I’m sure many readers have heard the arguments by some conventional analog manufacturers and CCTV integrators. “It’s too expensive” or “it’s too complicated” seem to be the most common. But in reality IP video is neither too expensive nor too complicated to begin replacing analog and DVR technology in stores today.  Let’s examine why.

First let’s start with the basics – IP video is cameras, cabling and recording, the same as CCTV. But the real difference between these technologies is freedom of choice. With IP cameras, users have a choice of standard resolution (which is the equivalent of 640 TVL analog) which is 20% greater than the highest resolution analog cameras. Most IP cameras also use Progressive Scan technology which captures motion in still frames significantly better than analog.

Images Courtesy of Axis Communications

Users can also select megapixel cameras. These fixed cameras allow images to be viewed live or stored and then when needed, zoomed in with no pixilation. The ability to forensically enhance video is an incredible technological advancement for retail loss prevention

Images Courtesy of Axis Communications

This outstanding picture quality also can be captured at speeds as high as 30 frames per second (real time).
The first question about capturing this much information must be “won’t it use a lot of bandwidth?”, a valid concern for video going across the wide area network or WAN. But most IP cameras have the ability to “dual stream” the video.  What is dual streaming? It’s intelligence inside the IP camera that allows it to send two separate video streams out. The first video stream can send a rich picture format of MJPEG at high fps to the storage device (more on that later)and the second stream can be configured to meet the requirements of the network.  This means the second IP video stream can be formatted in a network friendly format like mpeg4 (or the soon to be released H.264) at whatever fps, compression and resolution is needed to fit through your company’s available bandwidth. This enables asset protection to do live remote monitoring and real time investigations. The ability to control every bandwidth aspect of every camera is also why IP video is IT friendly. The retail IT group is far more likely to allow true IP video on their network. By contrast, most DVR’s offer only a global setting and proprietary compression technology. In many cases their throttling of bandwidth is either insufficient to control the risk of to the network, or is compressed to the point that analog video is severely degraded when viewed remotely. I’ve seen the GUI on some DVRs that are operating at such tiny resolution for remote viewing that video itself was in a 2” window. When it was enlarged, pixilation turned it into a blur.
I must add that IP video isn’t magic. If you only have is a 56K line going to the store, IP video won’t provide youtube type images. But IP video will allow you all the options available to shape the remote video to whatever your bandwidth will allow. 

Now that we know how we can stream our IP video across the network, the next logical question is how do we store that video to view forensically? This seems to be a question that stops many IP discussions before they get started. Remember that the DVR is basically a video server with a preconfigured hard drive? Well today retailers can build their own video servers and hard drive storage at a fraction of the cost of a DVR.  With open architecture video management software retailers can use common off the shelf components (COTS) like Dell, HP or IBM to construct their own storage solution. 

Retail IT department are already buying servers and storage for POS and Loss Prevention is finding they can leverage this buying power to build their storage solutions. For example, a grocery chain headquartered in Virginia recently built a 14 TB server for under $7,000. Those 14 terabytes of storage house the data for 44 megapixel cameras. This loss prevention director summed it up best when he said the quality of the video image was the single most important factor in moving to IP. With the money he saved from buying three DVRs and cabling savings, and power over the Ethernet (POE) he built a superior IP camera system for the same cost as a conventional analog camera system.  Additionally, he built his storage solution (Along with a little help from his IT department) to insure reliability. How? They simply allocated two of the 500GB hard drives as back-ups in case any of the hard drives failed.  This means if a hard drive did go down the video simply starts recording to a back up drive. No more lost evidence, ever. And by the way, since this server is just another server on his company network, the IT department does health monitoring and preventive maintenance on his hard drives.  If a hard drive ever does fail, his IT department can dispatch to the store and simply “hot swap” the failed drive without even shutting down the system. Add no system downtime to the list of IP benefits. And no repair bills or trip charges either. 

So far we’ve covered better video quality, remote connectivity and system reliability. But what about cost? We touched briefly on the experience of one Loss Prevention Director, but let’s dig a little deeper to help create a real world ROI. Before we do, I’ll insert a caveat – you can absolutely buy some low end analog CCTV components cheaper than IP video cameras. At this year’s ISC West show I saw 3.5mm mini-domes that had an MSRP of $129 and 16 channel tower DVRs under $2,300. But this brings up some sobering questions that need to be answered.

 How much value can you put on poor quality video? If the surveillance video isn’t sharp enough to identify a criminal what good is it? One major retailer in the US recently used IP video to identify the tattoo on the forearm of a masked armed robber. The tattoo, after the video was enhanced, was recognized by a tipster when shown on the evening news. That piece of IP video resulted in an arrest and confession. It’s difficult to put an ROI to that but there is certainly a value.

The next question to ponder is “what’s the life expectancy of these analog components?” It’s hard to say, but my industry experience has been that cheap fixed cameras usually burn in an image (ghosting) or fade out within a year.  DVRs, as mentioned earlier, typically last three years. If that’s an acceptable replacement cycle then don’t forget to add the labor costs of having them replaced. Labor usually includes trip charges, 1 hour on sight minimums and lift rental and that easily double the cost of the component itself. If it costs $500 to replace a $129 camera every  two or three years is that a better return for your money? The other consideration is system downtime. What’s the cost of not having surveillance on a given area? And how do you measure losing 30 day of video when a DVR goes bad? Again it’s tough to put dollar figures to these questions, but they can’t be ignored.
From a new store build perspective, IP cabling is 30-40% less expensive than analog cabling. CAT5/6 cable uses 60% less material than the combination of coaxial cable, 18/2 power wire and BNCs required for analog cameras.  The cost of running the CAT 5/6 cable is cheaper as well. IDF closets already in the store can accommodate the switches needed for IP cameras or encoders reducing the total length of cable needed and the associated labor to pull it.  With analog every camera is typically a home run back to security office. Another savings is that structured cabling doesn’t require low voltage wiring permits or licensed installers.

We also need to factor in scalability here. If you have 17 cameras you’re going to need 2 DVRs, even if the store only needs 20 total cameras. Additionally, the proprietary software of the DVR may not be able to remote connect to both of those units simultaneously – meaning you’ll have to disconnect from one to view other cameras in the same store. From a video storage standpoint we already examined the video quality concessions you may make to achieve 30 days of storage. By comparison 1 TB of storage on a COTS server now costs around $150. Imagine adding 10 TBS of storage for $1,500.

Organizational continuity is another soft coast that’s difficult to define.  When you change DVRS you change the software to operate them. Thanks to open architecture video management software (VMS), you’ll never go through the “learning curve” on a proprietary DVR GUI again.  When IP hardware is eventually replaced, the VMS will update to integrate ate it. You’re VMS will still have the same “feel” in 5 or 10 years.  The only difference is it will drive the latest hardware or software applications. An open architecture video management system will also allow your team the ability to integrate other applications like POS exception reporting, Data Mining, Fire Intrusion and access control alarms, EAS notifications and case management. From an investigation time standpoint how many hours could be saved having these tools integrated into your video platform?

In the big picture video technology has again reached a transition point for retail loss prevention. Is your organization ready to move out of the box to start benefitting from it now?