NIJ Journal No. 254 • July 2006

by Jeffrey P. Cohn

About the Author
Jeffrey P. Cohn is a freelance writer/reporter.

How can school administrators, teachers, staff, and parents make their schools safe for adults and children alike? How do you let parents and other authorized individuals into the building while keeping unauthorized people out without using up staff time to check identities and permissions? How do you know that a person entering a school building is who he or she claims to be? And how do schools resolve these questions without invading someone’s personal privacy?

One way involves a security system that links eye-scanning cameras with computers to identify people who have been preauthorized to enter the schools and then, once their identity is confirmed, lets them in by unlocking the door. The system has been adopted by three Plumsted Township schools in New Egypt, New Jersey, under a $293,000 science and technology grant from the National Institute of Justice. More recently, NIJ awarded a second grant to install a similar eye-scanning system in another, more demographically diverse New Jersey school.

In addition, NIJ funded an evaluation of the field test of the technology in the New Egypt schools. 21st Century Solutions, Inc. conducted an independent evaluation of the project, working in partnership with the schools and NIJ.

Nicknamed T-PASS (an acronym for Teacher- Parent Authorization Security System), the system in New Egypt identifies people using cameras that focus on 240 separate points on their irises. The iris is the round, pigmented area surrounding the pupil that controls how much light enters the eye. The experimental system represents the first use of iris recognition technology as a security measure for schools in the United States. Elsewhere, iris scanners are used to track inmate movements inside a dozen or so U.S. jails and to ensure that any prisoners being released are indeed the right ones. They are also used to identify some people entering Canada from the United States; some airline passengers at Reagan National Airport in Washington, DC, and other U.S. airports; and ATM users in Great Britain.

The ABCs of Biometrics

The use of iris scanners falls under what scientists and engineers call “biometrics.” Biometrics refers to a variety of computer-based technologies for recognizing individuals and verifying their identities using one or more of their physiological and/or behavioral characteristics. It has the distinct advantage of not requiring us to remember a user name, password, or series of numbers while confirming that we are who we claim to be. Biometrics is more reliable than traditional identifiers, such as driver’s licenses and identification or swipe cards, because it relies on individually unique characteristics. And because it is tied to a computer, biometrics is fast and provides a record that other methods usually do not.

Biometrics systems can use one or more of several different physical and/or behavioral characteristics for identification and verification. These include iris, retinal, and facial recognition; hand and finger geometry; fingerprint and voice identification; and dynamic signature. Some methods, like iris scans, are more technologically and commercially advanced than others. Which biometric method works best varies significantly from one application to another and even from one vendor to another. It depends on how and for what purpose the system is to be used; the level of accuracy and reliability required; and such factors as cost, speed, and user acceptance. None provides 100 percent accuracy.

Whatever method is used, biometrics basically involves a three-step process. First, a camera, scanner, or other sensor takes an image or picture. Second, that image is made into a pattern known as a biometric signature. Third, the biometric signature is converted into a mathematical pattern and stored in a computer. In iris recognition, the camera takes a picture of a person’s eyes. The image is fed into a computer, which compares that image with ones already in its files until it finds—or fails to find—a match.

Seeking Security in New Jersey

New Egypt is a small town in rural southern New Jersey about 45 miles east of Philadelphia. The school system has about 1,700 students in three schools—an elementary, middle, and a new high school. New Egypt school officials were unaware of biometrics in 2002 when they realized their schools needed a new security system. At the time, the schools used a swipe-card system that was aging and did not always work. Plus, there weren’t enough cards for everyone who needed one. School officials knew they had to improve not only the perception, but also the reality of school safety. They sought to develop a security system that would allay concerns and control access into the school buildings better than the swipe cards. They also wanted to use an innovative technology that could serve as a model for others.

After considering alternative biometric technologies, New Egypt officials chose iris recognition, one of the most reliable systems. Unfortunately for the school district, no complete iris scanning system existed that could be purchased and installed off the shelf. Instead, working with private vendors and NIJ, the school system developed its own iris recognition system.

New Egypt was able to buy 11 existing cameras, placing 6 inside and 5 outside the elementary school’s doors. Vendors had to write new software packages that would allow the cameras to send data images of scanned irises to a computer, tell the computer to search for a match, and then allow the computer to unlock the school doors once an individual’s identification was confirmed.

As the iris recognition system was being developed, school officials kept parents informed of the plans and encouraged them to participate in the voluntary program. All told, nearly all of the schools’ teachers and staff members and more than 700 elementary school parents had their eyes scanned into the system. The middle and high schools were not included in the test because far fewer of their students were taken out of class by parents or other family members during the school day.

A Passing Grade

For the most part, iris recognition worked. Of the more than 9,400 times someone attempted to enter the school using the iris scanners, there were no known false positives or other misidentifications. Indeed, the system provided an accurate identification and unlocked the door 78 percent of the time. Of the failed attempts, 6 percent resulted from people using the scanners who were not enrolled and thus whose iris scans were not in the computer. Another 16 percent were due to problems with outdoor lighting or someone not lining up his or her eyes properly for the camera to read accurately.

Most importantly, the iris recognition program seemed to make parents, teachers, and staff members feel safer in the school. When questioned as part of an outside evaluation of the program by 21st Century Solutions, parents who responded to the survey said at first they perceived little or no change in the efficiency of the sign-in process, the security problems within the school, or in the overall safety of the school neighborhood. Later, as people got used to the scanners, most parents said they believed the T-PASS system provided greater security than the previous swipe-card one and was easier to use than ringing a buzzer and waiting for someone to open the doors. They also reported being able to enter and leave the school much more quickly when picking up their children during school hours than were parents who continued to sign in and out manually.

Similarly, teachers and staff members at the elementary school told program evaluators that they perceived school security as significantly increased. They felt that problems such as outside people getting into the schools easily and staff members leaving doors propped open had declined. The elementary school secretaries, in particular, reported fewer parents walking around the school looking for their children.

Still Some Problems

Some problems with the new iris scanners and security system arose, as one would expect of any new technology. For example, during the first few days the cameras often froze up and would not work. Some felt that the signs telling people how to use the scanners (or the traditional buzzers for people who had not yet had their eyes scanned) were confusing. And as noted above, some people could not seem to line up their eyes properly so the cameras could accurately scan them.

The latter problem was particularly acute among older staff members and among people who have a dominant eye. It was partly overcome by advising people to try a second or even a third time. In some cases, school officials spent extra time showing people how to position their head so the camera could accurately read their iris. Schools in Freehold Borough, the next New Jersey district to test the iris recognition system under an 18-month, $350,000 NIJ grant, will use newer cameras that have two lenses rather than one. That will provide a more accurate reading even when people still cannot align their eyes properly.

A more serious problem was related to the use of outdoor cameras. Those cameras often failed to correctly identify people whose irises had been scanned, especially when they were in direct or bright sunlight. There were even problems accurately reading irises on gray, cloudy days. Most of the 16 percent failures noted above were due to sunlight affecting how the cameras could read the irises. In some cases, these problems could be overcome by placing a hood over the outside cameras to shield them from the sun.

Other problems had less to do with the technology or computers than with personal behavior. Many well-meaning students, teachers, and parents—once their irises had been scanned and the computer had unlocked the door for them—held the door open for another person entering the building behind them. Though the intent was good, the practice let others enter the school without having their irises scanned. Known as “tailgating,” the problem declined when school officials reminded teachers, staff, parents, and students not to hold the door open even if they knew the second entrant. Additionally, both the New Egypt and Freehold schools are installing laser beams emanating from the ceiling that will detect a second, unscanned individual attempting to enter the building behind someone else and sound a buzzer in the school office.

A similar problem involved teachers, staff members, and others who went outside the school on their lunch break or between classes to eat, smoke, or talk to their colleagues. Often, these individuals propped open a door behind them so they could get back into the building easily without going through the iris scanners again. School officials even found a brick placed by one door, used to prop it open. Again, the problem declined when officials reminded school employees and parents of the need to keep the doors closed and locked for security reasons.

Finally, before the iris recognition system was installed at the New Egypt elementary school, some parents expressed concern about privacy issues and the sharing of data among computer systems. To safeguard the personal information of parents and students using the system, the school recorded only the user’s name and driver’s license or other personal identification number. School officials promised users that their names and personal information would not be shared with any other data systems. Teachers and staff had their Social Security numbers and home addresses entered into the system, but that data represented information the school already had.

In the end, NIJ and New Egypt school officials concluded that the iris recognition experiment showed promise. As one school official put it: “The project helped build community pride. We were the first to do this. In 20 years, you’ll see biometrics in schools all over. All you have to do is look into a camera.”

The evaluators note, however, that there is little research on the overall effects of access control technologies on school safety. Most of the so-called “normal crimes”—minor thefts and assaults—that characterize daily life in American schools are committed by people who are supposed to be there. Because access control technologies such as the iris scanner are really targeted toward keeping out those who are not supposed to be in the building, the technologies’ impact on this type of crime is likely to be limited. And because outsiders constitute such a small minority of the people who commit crime in schools, the impact of these technologies might even be difficult to detect. So biometrics technologies such as the iris scanner should be considered as only one possible element in a school’s overall safety plan.

NCJ 214114

For More Information

• Uchida, C., E. Maguire, S. Solomon, and M. Gantley, Safe Kids, Safe Schools: Evaluating the Use of Iris Recognition Technology in New Egypt, New Jersey, final report submitted to the National Institute of Justice, Washington, DC: 21st Century Solutions, Inc., August 2004 (NCJ 208127), available at www.ncjrs.gov/pdffiles1/nij/grants/208127.pdf.