At first it was subtle. Reports were coming in that lights had been left on overnight and classrooms had been left unlocked. Then students and teachers noticed that school windows were being left open overnight and burned-out emergency lights were not being replaced. Eventually, PCs, laptops, stereos, and calculators began to disappear from the campus. Clearly, security officers at Escuela Campo Alegre (Happy Field School)—the American International School in Caracas, Venezuela—were skipping their overnight patrols.
Escuela Campo Alegre, established in 1937, is an English-language school located in the Venezuelan capital. It serves children from ages 3 to 18—preschool through high school. The school’s 600-plus students, many of whom are the progeny of diplomats and corporate executives, come from more than 40 countries around the world. The 115 educators that make up the staff hail from 12 different nations. About one-third of the students are U.S. citizens.
The various facilities, totaling 200,000 square feet, include not just classrooms, offices, and libraries but also a 650-seat theater, an Astro Turf sports field, and art, music, and drama centers, among other features. Because technology is a major enabler of learning at the school, the site includes expensive equipment that requires more security than might be needed for school supplies.
Among the learning tools to be protected from theft or vandalism are five centralized computer labs, video projectors in classrooms, at least one networked computer in each classroom, four mobile wireless “labs” equipped with wireless laptops, a multimedia lab for audio and visual production, and various other supporting and peripheral devices and components of a wired infrastructure.
Given the paramount value that the school places on student and staff safety, as well as the equipment dispersed throughout the property, security patrols are important. A nine-person contract security staff is charged with patrolling the entire four-and-a-half-acre property. The challenge for management is to ensure that security personnel carry out their patrol duties.
Management used to do this by having security officers mark items and locations on a checklist to indicate that they had been at each required stop along their patrol route and had done what was required of them. But over time it became clear that there were obvious discrepancies between what they had checked off and what had really been done. Management suspected that many officers were not conducting their patrols, even though guards were not shirking their duties within the view of CCTV cameras. The school sought a better way to verify performance.
A New System
During winter break in 2003-2004, the school brought me in as the new security manager. After seeing the problem, I decided to switch to an electronic guard tour system, which the school purchased for about $2,500. The system consists of several eight-ounce “data probes,” or guard-tour wands, loaded with special software, as well as 80 “buttons”—chips enclosed in steel cases—that can be affixed anywhere on the campus. Touching a wand to a button enables the wand to read a unique serial number encoded on each button, recording the time and date of the event. That information is stored in the wand and can be downloaded to a computer. Buttons were affixed at 80 stations throughout the campus.
The system includes a “master button” that, when touched by a wand, indicates to supervisors through different tones whether the wand has been used to carry out a full tour of buttons. Also included in the system are a cable to download the data and a leather carrying case to hold the equipment when it’s not in use.
After the system was installed, we established procedures for use. At the start of each overnight shift—between 8 p.m. and 10 p.m., depending on whether any sporting or other events were occurring at the school—each officer would sign for his personal guard-tour wand. When the wand was returned at the end of the shift, a supervisor would check to make sure that it was still functioning properly.
Officers would carry out their rounds by placing the instrument near each of the buttons affixed to the stations. That process documented that the officers had at least physically visited each station. Each time the instrument was placed near a button, the system recorded the location, day and time, and name of the guard. Early the next morning, when I arrived at the school, I would download the information from the wands to a computer. The data were converted into reports that detailed the activity.
For the purposes of guard shift work, we divided the school grounds into four sections: the school’s perimeter, the guardhouse at the entrance to the property, the main academic building, and a perimeter around the main building. Each shift, during which all four guards were on duty, was broken into quarters, or two-hour periods.
During each quarter, a different guard would serve as the leader. The leader remained at the guardhouse, watching CCTV feeds from all around the campus and staying in touch with the other officers via walkie-talkie. To ensure that the leader didn’t doze off or wander away during those two hours, he was required to press a button in the guardhouse every ten minutes.
The other three officers were responsible for their rounds in one of the school’s other three sections. They were encouraged to walk slowly and keep their ears open as well as their eyes. The school is surrounded by a forest, and animals such as possums, squirrels, and dogs, often roam onto campus. The animals themselves are not really a threat, but they can be a distraction. Guards are expected to learn to distinguish between stray animals and human intruders by the sounds they make.
The three guards were directed to periodically report to the leader to indicate where they were and if anything seemed abnormal or noteworthy. They also had to place their wand to each button they passed to ensure that they walked their route. (The 80 buttons are divided at roughly equal distances throughout the sections.)
At first, each guard was required to patrol continuously for about an hour, then remain stationary (or “static”) for 15 minutes, then walk for another 45 minutes. The wand would beep when it was time to change.
At the two-hour mark, another guard would become the leader and the other guards switched sections of the school grounds. This rotation was designed to keep guards from getting bored and to ensure that the routes were continually patrolled by a new set of eyes. In this way, one guard might observe aberrations that other guards had overlooked.
To make sure that the guards not only patrolled when they were supposed to but also stayed at their post when they were supposed to be stationary, they were assigned places that were in the field of view of CCTV cameras. Both the officer in the guardhouse and I had access to these feeds.
Tweaking the System
After several months, patterns developed. The reports showed what behavior was typical of specific guards, such as time required to go from station to station depending on the day of the week, time of day, and section of the campus. Other behaviors showed up as inconsistencies or aberrations: stations being skipped, inordinate time between stations, missed rotations, attempts to remove the batteries from the guard-tour wand, or attempts to short-circuit other components of the guard tour equipment, for example.
Six months after the system was installed, it became apparent that the level of turnover among the night shift was unusually high and that the performance of some of the remaining staff had flagged. After speaking with some of the staff, we concluded that the guards were getting exhausted walking about six hours per night and only getting one day off per week. At first, we tried to address the problem by increasing guards’ static time to 30 minutes. As a result, guards would patrol for a half hour, then remain stationary an equal amount of time. The guard tour system was reprogrammed to accommodate this change.
Three months later, guards were still complaining about the amount of walking, so we reprogrammed the system to increase static time to 45 minutes. This schedule—walk for 30 minutes, stay at a post for 45—is proving best for everybody. Because 30 and 45-minute rotations don’t fit neatly into a two-hour period, each guard now stays in a section for two and a half hours (30 minutes walking, 45 rest, 30 walking, 45 rest) before switching to another section.
We also realized that the schedule of three days on, one off was contributing to guard exhaustion and playing havoc with their personal lives. Schedules have been changed so that each officer works as follows: a day shift on the first day, a day shift on the second day, a night shift on the third day, a night shift on the fourth day, then two days off. Because of the timing of the shifts, guards now get nearly three days off to spend as they choose.
Another concern involved having to push the button at the gatehouse every ten minutes, a task that some found to be monotonous. We were able to sell the officers on its importance, however, by stressing that pushing the button provided evidence that the officer was in the guardhouse when a theft or other incident occurred elsewhere on campus.
As more time passed, security noticed that most incidents occurred in the main building. We decided to lock down the building overnight and eliminate the building patrol. Only the supervisor would have the key to the building. We were able to convince the administration to let us install CCTV in all the corridors, 16 cameras in all, as long as none were installed in classrooms or offices.
A new evening-hours procedure for access to the building was implemented as well. After 5:30 p.m., all entry and exit occurs only through the building’s main door, which is presided over by a guard. The building is emptied and locked for the night at 8 p.m.
Evaluating Guard Service
Realizing that some officers simply weren’t suited to the demands of the job, we decided to take a closer look at the guards’ complaints and analyze the turnover rate. At the 18-month anniversary, about 230 guards had come and gone. We reached the following conclusions about the program.
First, older and overweight guards struggled with the walking regimen. We now ask the guard contractor to supply the school with officers who are fit.
Second, guards without some level of higher education failed to learn the system well and couldn’t understand what the security department was trying to achieve. Now we seek candidates who have at least a high school diploma.
Finally, the best guards turned out to be those who began their careers with us. Guards with experience at other sites came with too many bad habits and proved difficult to retrain.
Given the guard provider’s poor track record of providing suitable officers, school security decided to reevaluate the contract. We sought out a company that emphasized professionalism and physical fitness.
While we demand good quality from the provider, we also do our part by appropriately compensating and appreciating officers in ways both large and small. We used to provide three meals a day, plus a midnight snack, for all officers, depending on which shift they were working; we now provide bonus money in lieu of meals. We also pay bonuses to guards who fulfill their duties as required, evidence of which comes via both observation and guard-tour reports. An Officer of the Month designation is awarded to those who exhibit superior performance. The prize is a gift certificate for dinner for two at a restaurant.
Since we changed guard providers and implemented the guard-tour system, vandalism has decreased drastically, though exact numbers aren’t available. Only two significant cases have occurred while the new tour system has been in place, and both involved the previous guard company. Both were resolved through the use of the tour system.
In February 2005, a teacher reported that someone had been using her computer overnight to view pornography. Security began the investigation by asking the technology department to identify the computer and the time that the culprit used it. We took that information and cross-referenced it with data from the tour system. By looking at which guards were on duty at which locations and at which times that night, we immediately narrowed the list of suspects to one guard, who turned out to be the offender.
In May 2005 a series of three thefts occurred from different parts of the campus, all overnight. Reviewing our reports, we narrowed the list of suspects to the few who were near the facility at the time and on the day when a theft took place. One guard in particular emerged as a suspect, so we examined his records for those nights.
In the areas where the thefts occurred, the average time between touches of wand to button was 15 to 20 seconds. It turns out that the suspect was taking five or six minutes between touches. In addition, he was not stationing himself in view of a camera during his static periods. Since prosecuting a criminal case would have required a lot more evidence and time, we simply asked the guard service provider not to send that officer to our property again.
Replacing the patrol of the main building with CCTV has also been beneficial, improving security and yielding some savings. The CCTV system, which cost $12,000 to purchase and install, eliminated the need for one guard, a savings of 25 percent on officer costs. During the course of a year, that translates into more than $8,000. Consequently, the CCTV installation paid for itself within 18 months.
The effect on officers is appreciable too. The new scheduling gives them time to spend with their families and to attend to personal matters. The change has improved morale.
The system isn’t foolproof. Minor incidents still occasionally occur—windows are left open, doors remain unlocked, faulty lighting may go a day or two without repair or maintenance. But between the new system and the improved staff, campus security is getting higher marks than ever.