Terror Hunt: Hounding With Biometrics

10 Apr, 2002    ·   718

Maj Gen Yashwant Deva argues for the use of Biometrics in terror hunt


They will be brought to justice or justice will be brought to them

 

- George W Bush

 

How do we recognize Osama bin Laden or Masood Azhar to bring them to book? Does the answer lie in biometrics, the science of recognizing a person using distinguishing traits? Physiological and behavioural characteristics are unique to an individual and can be scanned, and scaled and scrutinized.  The former include fingerprints, retinal and iris peculiarities, hand geometry, voice patterns, and facial recognition; the latter encompass gait, speech, handwriting, signatures, mannerisms, gestures; and traits that remain to be explored. Of greater scientific precision is the DNA test, of significance to verify the true identity of an individual, when boarding an aircraft, entering a high security zone, immigration control, police verification of suspects, security identification and collecting evidence to bring terrorists to justice. Passports, identity cards, personal identification numbers (PINs), photo identification and “tokens” are undoubtedly a help, but can be compromised. On the other hand, a digital biometric reader or identifier is distinctly more precise and cannot be easily duped. Sensor fusion is a natural cognitive attribute of the cerebrum and, if imitated artificially, can lead to functional integration of gadgetry. 

 

 

Biometrics was a developed science during ancient times in India . References abound in the Puranas and Mahabharata of identification by physiological parameters. Elsewhere, too, there are records of it being practiced, though more as a ‘black art’ by some high priests. Later, in the nineteenth century, there was a marked spurt in research. During the days of the Raj, the science of criminology worked on the mistaken premise that physical features and characteristics were associated with criminal tendencies. This led to a variety of measuring devices and collection of data to detect latent tendencies that had the potential for legal abuse. However the idea of measuring individual physical characteristics was accepted. With the parallel development of fingerprinting, biometrics acquired international acceptance. Records of complexion, eye colour, height and so on were routinely kept to augment fingerprints and photographs.

 

 

Today biometrics is a frontier technology. Over the last decade the biometric industry has come a long way. Many devices are in the market, and some are reportedly used by hi-tech criminals. 

 

 

Biometrics has many versatile but yet-to-mature applications. Of significance here is detection of bank frauds, Automatic Teller Machine (ATM) operations, workstation and network access, e-business transactions over the Internet, biometrics embedded credit card, key encryption security enhancement, digital water-marking, public identity smart cards and, above all, voice recognition for telephonic conversations. Biometrics can lend greater potency to digital signatures, which the country has adopted from Republic Day this year.  The sky is the limit for possibilities that can be put to use for cyber intelligence operations, e.g. tracking hawala (money laundering) and illicit drug trails, monitoring Internet Relay Chats (IRC), e-mails, steganography and for counter-hacking and counter-terror operations. 

 

 

There are undoubtedly many problems and gray areas in realizing the full potential of this technology, some technical, some commercial, but more psychological. The first problem lies in obtaining the sample or biometric template of the individual, whose identity is to be verified.   Undoubtedly this is a Herculean task and often a sheer impossibility. However it would be folly if biometric samples have not been gathered and archived for example in respect of Osama bin Laden or Masood Azhar. The technology can work in controlled lab conditions or perhaps in a virtual environment, but not for real-time or large scale. The unit cost of gadgets is exorbitant and currently there is no demand for mass application. 

 

 

The performance of a biometric device is measured in terms of its failure rate—“accept-fail” or the likelihood that an impostor may be accepted, and “reject-fail” or the likelihood of rejection of a bona fide person. The bias for either leaves scope for manipulation and spoofing, besides putting a question mark on its confidence level. Take for instance fingerprints, the absolute uniqueness of which is often debated. Recently a justice in the US has rejected the admissibility of fingerprint evidence, terming it subjective. Voice verification products also suffer because of variability of both transducers and environmental acoustics. Odour identification is an exciting domain of application, but is yet to acquire a viable character. 

 

 

Much work continues to be undertaken and it will be interesting to watch its progress. Integration, scan-fusion and networking will achieve the desired quality and product reliability. Biometrics is a transition from science fiction to an over-the-horizon technology. Let us create public awareness about it and get down to some serious R&D. 

 

 

 

 

 

POPULAR COMMENTARIES