Skip to content

Watching Over Guests

THE INTELLIGENT-VIDEO management system had only been installed for a few weeks in Chicago’s boutique Talbott Hotel when its worth was tested. The hotel’s night auditor used the system’s camera images to track the path of a suspected prostitute who had allegedly broken into a room and stolen a Neiman Marcus bag and a purse. When the police arrived, the hotel’s security staff was able to quickly furnish them with video of the incident, burned onto a CD.

The system from 3VR Security, Inc., had captured a clear shot of the suspect’s face. Within a few hours, she had been found on the streets and apprehended, and the belongings of the robbed hotel guests were returned.

Troy Strand, the Talbott’s general manager, decided to purchase a digital video management system after becoming frustrated with his previous digital equipment.

The old system required sitting “for hours and hours, holding your finger on the mouse, scrolling through video” trying to find particular images, he explains. It also took four hours of expensive staff training time, and fell short of delivering sundry promised features, including facial recognition.

In contrast, the 3VR system requires employee training of about 15 minutes, and it has generally performed as promised. For example, to aid police on the case involving the robbery by the prostitute, the Talbott’s security staff used two key components of the 3VR system: its ability to rapidly narrow down video footage based on the hour, the location, and any unusual physical movement; and its ability to help staff find suspects using the facial recognition feature.

A bundled-software-based appliance, the 3VR consists of servers that look like DVRs and are mounted in a rack. Also part of the package is software that can be installed on a desktop computer. Cameras are purchased separately.

The 3VR software is accessed either by connecting a computer monitor directly to the server or by using the CD to install the client program on a computer, thus connecting to the system via the network. The number of servers depends on the size of the system: Strand’s setup has about five. 3VR personnel do the installation, which at the Talbott took one day.

The 3VR software works with analog cameras, meaning that Strand could still use his existing 32 cameras and only had to buy new cameras to the extent that he wanted to extend the system’s coverage to other areas of the hotel property. That helped hold down the cost of the upgrade.

Overall, Strand says a user can expect to spend about $2,000 per camera port, excluding the price of the cameras themselves.

The Talbott’s 3VR system is wired into six PCs belonging to managerial and security staff. The ability to use the system on several PCs is included with the initial 3VR package, but there is a charge for additional connections.

The package of software consists of various modules. Each module consists of a set of capabilities, such as motion detection or facial recognition, which is programmed into the system’s memory card.

Additional features can be installed simply by upgrading the memory card, which sits in the front of the 3VR appliances.

With the motion detection module, for example, if a camera will be trained on a basement where there is usually no activity, the system can be programmed through a series of configuration screens so that the basement can be recorded continuously, or only when motion occurs.

Live video can be viewed via 16 panels on one screen. Thus, 16 of the cameras can be viewed at once, or various views from the same camera, such as zoomed and not zoomed, can be seen simultaneously. If a panel of video is deemed important, it can be e-mailed—either as a snapshot or a brief video clip—or copied to a CD.

The system can be programmed so that cameras send alarms according to criteria such as unusual movement. The end user defines what constitutes unusual. For instance, if someone goes through a particular back door of the Talbott between 11 p.m. and 7 a.m., the system has been set to alarm.

The facial recognition feature makes it possible for security to set the system to look for persons on a watch list. For example, if a previous employee or guest who was banned from the premises is spotted by a camera, it triggers an alarm. The emergency message is transmitted via e-mail, by cellular phone, visually on screen, or audibly through an alarm relay. Strand says false alarms have largely been prevented by feeding the system at least 10 distinct photographs of individuals whom he wants recognized.

There is also the ability to archive data. The Talbott currently has three terabytes of space, with an option to expand. That works out to about six months’ worth of video.

It is easy to go into archived data via specified search parameters, such as time or type of event. For example, if a specific incident occurred in the hotel lobby at a designated hour, the video search can home in to that place and time. “The search technology is quick,” says Strand. “You can go to a time, select as many cameras as you want, click on the tabs, and see what you need in seconds.”

The word about the system’s effectiveness has spread, acting as a deterrent to in-house theft as well. “We had employees steal and not get caught with the initial system,” says Strand. “Now employees are so aware we are watching, they don’t attempt anything.”

Strand’s only complaint about the 3VR is that it only works with monitors that are at least 20 inches; smaller screens can’t accommodate all the information. The Talbott’s existing monitors are mostly 17 inches and 15 inches, and the hotel didn’t want to have to replace large numbers of them. So the hotel bought a few 20-inch monitors while 3VR works on scaling its software to work on smaller screens.

By that time, Strand expects to be upgrading to a new system that includes new capabilities. One adds a license-plate recognition feature for the valet parking staff that will make it possible for cameras to zoom in on, read, and record license plates. That will allow security to use that information to determine who owns the car, if necessary.

Another capability is the “abandoned or removed object alert” recognition and alarm capability, which gives the system the ability to alert if an item is removed from a customary spot or if a foreign object is left in an area for more than a short time. The latter is especially handy in preventing hotel patrons from leaving their bags behind in the lobby when they depart the premises. It can also help to spot suspicious packages.

“We want to make sure we can protect our guests, our employees, and our assets,” says Strand.

(For more information: Tim Ross, executive vice president, 3VR; phone: 415/513-4601; e-mail:[email protected])

By Robert Elliott, assistant editor at Security Management.