Handheld biometric technology integrates the mobility and versatility of a video camera equipped with recording audio recognition and facial capture technology. Many of these devices are also equipped with a fingerprint reader and iris scanner.
While on patrol, an officer observes a driver roll through a stop sign. The officer activates her emergency lights and stops the vehicle. After approaching the stopped vehicle, the officer asks the driver for his license and registration, but the driver is uncooperative and refuses. The officer detains the driver and reaches for a handheld biometric recognition device attached to her police duty belt, raises it to the suspect’s face and presses a single scanning button. A “hit” on the device’s screen indicates that the driver is wanted for murder. Aided by innovative biometric technology, the subject is arrested. This is just one example of a potential application of handheld biometric devices for law enforcement.
While not yet widely deployed, these devices hold a lot of potential for law enforcement applications.
After providing a brief history of biometrics, this article will focus on handheld biometric devices, including their advantages and limitations, officer safety and legal concerns, and recommendations.
History of Biometrics
Biometrics, the development of automated human identification, had its origins around 6000 B.C. when the Assyrians, Babylonians, Japanese, and Chinese first recorded fingerprints. However, it was only in the late twentieth century that biometrics expanded to include many other forms of identification, such as retinal scanning, iris recognition, improved hand geometry, and more. Although automated facial recognition originated during the 1960s, it remains a relatively new concept. Realtime automated facial recognition systems were developed by 1991.
In March of 2011, the FBI disseminated information about a new Biometric Database, the Next Generation Identification (NGI), a Lockheed Martin program which steps beyond the Integrated Automated Fingerprint Identification System (IAFIS). While not yet completed, the NGI will provide “automated fingerprint and latent search capabilities, electronic image storage, and electronic exchange of fingerprints to more than 18,000 law enforcement agencies and other authorized criminal justice partners 24/7.”
The first Biometric Standards for the interchangeability and interoperability between biometric technologies, BioAPI Version 1.1, were approved through the InterNational Committee for Information Technology Standards (INCITS) and adopted by the American National Standards
Institute (ANSI) in 2002. The ANSI/INCITS (M1) 385-2004 and ISO SC-37 19794-5 Face Recognition Data Interchange Format serve as the current facial recognition standards. Since then, countless products and numerous combinations of biometric technologies have emerged.
The FBI created the Biometric Interoperability Program, approved by the Department of Justice in 2008, promoting “biometric-based information sharing between the FBI’s Integrated Automated Fingerprint Identification System and other federal, state, and international biometric systems.” Facial recognition joined with the largest DNA repository in the world, the National DNA Index System (NDIS), the COmbined DNA Index System (CODIS) and the Biometric Center Of Excellence (BCOE), and spurred the formation of the Facial Identification Scientific Working Group (FISWG). Facial recognition/identification capabilities became a BCOE initiative: It “involves the automated searching of a facial image in a computer database, typically resulting in a group of facial images ranked by similarity.” The two predominant approaches, “geometric (feature-based) and photometric (view-based),” calculate distances and ratios of features to a common reference point. Facial recognition is now used alongside iris recognition and fingerprint identification, but is not yet considered as accurate as the latter two. Error rates range from about 3.8% to 13%, depending on the facial recognition algorithm used. In comparison, fingerprint expert Simon Cole estimates that error rates for fingerprint identification range from 0.2% to 2.5%.
Versatility of biometric technology has increased and circulated to other fields, including social networking. For instance, social networking sites have utilized primitive facial recognition technology in profile linked photos. In addition, pattern recognition programs have been used on Web sites which search images and videos for a particular face.
Handheld biometric readers are capable of accessing multiple databases to identify and provide information about individuals who are identified through biometric characteristics. They have the memory capacity to store between one and five hundred thousand templates and to access remote databases, depending upon communication links.
In addition, GPS transmission capabilities of most handheld biometric readers may support 3G/ 4G cellular, Wi-Fi, tactical radio, and WiMAX or equivalents. Advances also allow the technology, including camera functions, to work in complete darkness. Depending on the size of the device, it may be easily stored on the person, such as on a duty belt, in cargo pockets or within hand baggage, but larger styles may have to be stored within the police vehicle.
Handheld biometric readers range from the size of a cell phone to a toaster, weighing between three and four pounds (about two kilograms). The handheld readers recently issued to sheriff’s deputies in Pinal County, Arizona, are only slightly larger than a pack of cigarettes and attach to an iPhone or similar device. There are, in fact, biometric applications available for cellular devices.
Advantages and Applications
Biometric information can be processed in less than a second. It can even differentiate between identical twins. Accurate readings are confirmed with up to 40,000 data points used in matching. The effect of glasses and other facial coverings on its accuracy remains unclear. However, contact lenses do not prevent an accurate iris scan reading.
The handheld reader poses no potential harm through exposure to the radiation and lights needed for its capabilities. Exposure to retinal blue light limits, ultraviolet radiation, infrared radiation, and thermal hazards is below the defined hazardous limit. The technology withstands temperatures ranging from 35 to 120 degrees Fahrenheit (0-50 degrees Celsius) and is operable between 10-95% humidity. The potential uses of the biometric devices are diverse, with applications extending beyond street patrol. For instance, the reader could be used in criminal investigations, such as identifying suspects or victims at crime scenes; in the search for missing children; and in Alcoholic Beverage Control (ABC) investigations of underage drinking. For example, if the identification of a homicide victim is in question, handheld biometric technology has the potential to determine identification from the iris, face, or fingerprints of the victim, saving investigators valuable time.
In the future, the biometric reader may potentially aid investigators in quickly examining suspects found at, or in flight from, the scene of a crime. The device may also be able to compare prints left at a crime scene with those of subjects in a fingerprint database.
Cases of missing children continue to be a public concern, especially as offenders can easily cross jurisdictional boundaries. Handheld biometric technology can assist officers in confirming the identification of known suspects believed to be involved in an abduction, even if from another jurisdiction. And, if an officer suspects that a child has been abducted or recognizes a recent abductee, this technology can help confirm the child’s identity by accessing databases of missing children nationwide. Whether an officer suspects that a child has been abducted or is simply lost, handheld biometric technology has the potential to quickly help to verify identifications.
The production and use of false identification (e.g., fake photo IDs) has proved problematic. Advanced computer graphics technology has made the production of false identification easier and law enforcement detection of it more challenging. For instance, juveniles are exposed to a plethora of options to conceal their identity, including borrowing others’ identification cards, purchasing fake identifications, and defacing otherwise valid identifications. If an officer is suspicious of the provided identification, it is possible that valid identification may be determined with handheld biometric technology.
There are some limitations to handheld biometric technology, including database limitations; storage, cost, and environmental restrictions; officer safety; and potential questions which must be considered by agency legal advisors. First and foremost, handheld biometric devices can only accomplish the benefits described earlier if they are linked to information in the appropriate databases. Furthermore, successful identification of an individual is contingent upon that person’s information being included in the database(s) to which the device is linked.
While the technology is portable, devices possessing iris, facial, and fingerprint identification may be too bulky for the duty belt. As noted earlier, however, they may be stored within a police vehicle. This compromises the mobility of the device. Mobility is also hindered by weather conditions. While the handheld technology can withstand high levels of humidity, its use is not recommended under precipitation (rain, sleet, snow, etc.).
Handheld biometric readers are expensive. Styles that only employ fingerprint technology are less costly and range from under one hundred dollars to several hundred dollars. For the biometric technology which includes fingerprint, iris, and facial recognition, the cost could reach several thousand dollars per device, not including database access costs. For instance, the Pinal County Sheriff’s Department recently purchased 75 readers which function as an iPhone attachment. The cost per unit is approximately $2,900 and the software package for the devices will cost another $9,000. The grand total will be higher, however, when the readers are implemented, as the reoccurring data package plan needed to support the technology costs approximately $40 per device. The Pinal County Sheriff’s Office is waiting on the implementation plan before a time frame can be set for training. Training will add additional costs and may be time-consuming, but it is vital to proper implementation. Smaller agencies or those on limited budgets may find the devices cost-prohibitive. However, agencies wishing to purchase the handheld biometric devices could possibly seek external funding and grants.
Depending on the police agency, department duty shifts often range between eight and 12 hours. Since handheld biometric readers are operable for up to eight hours after charging, this could limit their use. Charging takes approximately three hours and can be completed through the charging port in a vehicle. For the reasons described above, use of the device may be limited on foot patrol, mounted patrol, and other patrol settings outside of a vehicle.
Another limitation is that the subject must be within close range, depending on the feature employed (fingerprinting, iris scan, or facial scan), for effective use. For instance, iris scan and fingerprinting require direct contact of the subject’s face and fingers to the device. The officer must be in close proximity (three to six feet) with the subject for facial recognition to function properly, thereby posing a potential risk for officer safety.
In addition, both the subject and user must be stationary. When the device is calculating identification and the officer reads the results, he (or she) must remain still and focused on the device which takes the officer’s attention away from the subject. In an article listing the top risk factors for officer fatalities, not keeping eyes on a suspect’s hands at
all times and taking a position in close proximity to the subject are noted within the top ten. It is therefore recommended that the device be used only when the suspect is in custody and properly detained. Backup may be required, thereby removing another officer from his/her assigned area. Use of the device may be limited in cases when backup is not available.
Legal and Privacy Concerns
Questions to be considered by police legal advisors might include: Do handheld biometric devices pose a privacy concern? Is evidence from handheld devices admissible in court? Does the use of biometric recognition technology constitute a search under the Fourth Amendment?
Handheld biometric technology may raise privacy concerns.Though biometric technology retrieves information in public settings, the information that this technology accesses may be considered private. Biometric readers can access identification information which is more revealing than a simple ID card. For example, readers scan biometric personal physical identifiers which are unique to the individual, potentially linking information within a database which can provide the driving record, criminal record, and Social Security number of the subject.
Another question for consideration is the admissibility in court of information from handheld biometric devices. However, the handheld biometric technology arguably would satisfy both the Frye and Daubert Standards. Frye ruled that expert testimony is admissible only when it has achieved a general acceptance in the relevant scientific community. The Daubert standard allows the court to determine admissibility for scientific or expert testimony based on how it has been evaluated in the scientific community, requiring that “the evidence must be based on reliable principles and scientific methods and the expert witness must be shown to have applied the scientific principles properly.” The Daubert standard, used in federal courts and most state jurisdictions, followed Frye on the assessment and admissibility of scientific expert testimony on technology. Biometric readers would likely meet all Daubert criteria: The technique has been tested; has been subjected to peer review and publication; potential error rates are decreasing; there are existing standards of maintenance and control; and the device has attracted widespread acceptance within the relevant scientific community.
Because privacy issues are a concern, it is important to raise the question as to whether the use of biometric readers constitutes a search under the Fourth Amendment. The Fourth Amendment indicates that a search occurs when a reasonable expectation of privacy, as recognized by society, is infringed by agents or employees of the government.
In United States v. Dionisio, the Supreme Court ruled that an individual’s expectation of privacy does not include characteristics which are constantly exposed to the public and easily under surveillance. Videos and pictures retrieved from surveillance recordings are often used for identification purposes. In this sense, the biometric reader is no different than proximate surveillance. But, is information retrievable by handheld biometric recognition private? Currently, the answer remains unclear.
In Hiibel v. Sixth Judicial District, the Supreme Court ruled against a defendant who was arrested after refusing to provide officers his name when asked (there was a law in the defendant’s jurisdiction requiring such identification to be provided upon request by the police, when the police had reasonable suspicion of illegal activity). The Court rejected the claim that such requirements for providing identification violated either the Fourth or Fifth Amendments.
In Brown v. Texas, however, the Supreme Court ruled that persons may not be detained for the purpose of determining identification when reasonable suspicion is lacking. According to Brown, reasonable suspicion is required to detain a suspect for his/her identification. These decisions may have implications for the use of handheld biometric scanners. At the same time, it remains to be determined whether use of the scanners requires detention of the subject.
The Supreme Court recently decided a case on warrantless GPS tracking of automobiles and its Fourth Amendment implications. The issue may be of interest for biometrics, in terms of implications for the electronic measurement of publicly observable information. The Supreme Court held that such tracking constitutes a search and, therefore, cannot be conducted without a warrant. The Court’s ruling appeared, however, to hinge specifically upon the definition of a vehicle as an “effect” entitled to Fourth Amendment protection and its implications for acquisition of biometric data remain to be seen.
While the Supreme Court remains the ultimate arbiter of privacy jurisprudence, biometric privacy cases have not specifically been addressed. For this reason, users of new technologies should give careful consideration to developing appropriate guidelines for their use, as noted below.
Recommendations and Conclusions
Agencies currently using or planning to use handheld biometric devices should carefully research the technical capabilities and requirements for the devices and then develop clear policies for their application. At a minimum, these policies should specify the purposes or situations in which the devices should be deployed and any guidelines or restrictions on their use.
Additionally, training is strongly recommended for officers to ensure the most effective use of these devices. The Institute of Electrical and Electronics Engineers (IEEE) Certified Biometrics Professional (CBP) program offers courses which set a baseline of biometric knowledge for those who plan to use the technology. Some agencies choose to adopt these courses as certification standards. Those who successfully complete the examination following the courses have demonstrated a level of proficiency in the field and are expected to use biometric technology in a competent and effective manner. However, these standards have not been accepted nationwide and, at the time of this writing, serve only as a guideline.
While the limitations have been noted, handheld biometric technology holds potential value for law enforcement agencies, including the facilitation of quick and valid identifications when personal information is lacking or suspicious. Handheld biometric technology is a recent development and, like all new technology, is inevitably prone to practical and legal review. However, its potential for law enforcement, as defined by its range of applications, appears extensive and promising. The handheld biometric reader does not replace effective law enforcement; it complements it. Biometric readers have evolved as yet another investigative tool for law enforcement officers.
About the Authors: Ms. Katherine Rasiak is an Honor’s Academy student at Radford University in Radford, Virginia. Ms. Rasiak can be contacted at firstname.lastname@example.org.
Dr. Tod W. Burke, a former police officer, serves as a professor of criminal justice and as associate dean for the College of Humanities and Behavioral Sciences at Radford University in Radford, Virginia. Dr. Burke can be contacted at tburke@ radford.edu.
Dr. Stephen Owen is a professor of criminal justice at Radford University in Radford, Virginia. Dr. Owen can be contacted at ssowen@ radford.edu.
This article is a contribution from Police & Security News. P&SN is bi-monthly law enforcement magazine that is packed with training articles and gear reviews for the patrol officer. P&SN is a valued supporter of BlueSheepdog and the Blue Crew. You can obtain a free subscription to the Police & Security News magazine by joining the Blue Crew.