Home­land Se­cu­rity sniffs out ID by odor; does it pass the smell test?

The Washington Times Weekly - - National Security - BY SHAUN WATER­MAN

The Depart­ment of Home­land Se­cu­rity plans to study the pos­si­bil­ity that hu­man body odor might be used to de­ter­mine when peo­ple are ly­ing, or to iden­tify in­di­vid­u­als in the same way that fin­ger­prints can.

In a fed­eral pro­cure­ment doc­u­ment posted March 6 on the Web, the depart­ment’s Sci­ence and Tech­nol­ogy Direc­torate says it will con­duct an “out­sourced, proof-of-prin­ci­ple study to de­ter­mine if hu­man odor sig­na­tures can serve as an in­di­ca­tor of de­cep­tion. [. . .] As a secondary goal, this study will ex­am­ine [. . .] hu­man odor sam­ples for ev­i­dence to sup­port the the­ory that an in­di­vid­ual can be iden­ti­fied by that in­di­vid­ual’s odor sig­na­ture.”

Of­fi­cials said the work was at a very early stage, but the an­nounce­ment brought crit­i­cism. Barry Stein­hardt, of the Amer­i­can Civil Lib­er­ties Union’s Tech­nol­ogy and Lib­erty project, said the plan showed that the depart­ment had “mis­placed pri­or­i­ties.”

“The his­tory of DHS’ de­ploy­ment of th­ese tech­nolo­gies has been one colos­sal fail­ure af­ter an­other. There is no lie de­tec­tor. This re­search has been a long, me­an­der­ing jour­ney which has taken us down one blind al­ley af­ter an­other.”

Mr. Stein­hardt said that even well-es­tab­lished bio­met­ric iden­tity tech­nolo­gies such as fin­ger­print­ing had re­sulted in in­ac­cu­rate iden­ti­fi­ca­tions of peo­ple, in­clud­ing Ore­gon lawyer Bran­don May­field, who re­ceived an apol­ogy from the FBI af­ter be­ing wrong­fully ac­cused of hav­ing had a hand in the 2004 Madrid train bomb­ings.

“None of the bio­met­rics for iden­tity have worked very well, with the pos­si­ble ex­cep­tion of

“This re­search has the po­ten­tial for en­hanc­ing our abil­ity to de­tect in­di­vid­u­als with harm­ful in­tent,” the no­tice says. “A pos­i­tive re­sult from this proof-of-prin­ci­ple study would pro­vide ev­i­dence that hu­man odor is a use­ful in­di­ca­tor for cer­tain hu­man be­hav­iors and, in ad­di­tion, that it may be used as a bio­met­ric iden­ti­fier.”

Amy Kudwa, a Home­land Se­cu­rity spokes­woman, said “proofof-con­cept” work was the ear­li­est

The pro­cure­ment no­tice says the depart­ment is al­ready “con­duct­ing ex­per­i­ments in de­cep­tive be­hav­ior and col­lect­ing hu­man odor sam­ples.” The re­search it hopes to fund “will con­sist pri­mar­ily of the anal­y­sis and study of the hu­man odor sam­ples col­lected to de­ter­mine if a de­cep­tion in­di­ca­tor can be found.”

DNA,” he said, adding that even fin­ger­print ev­i­dence was “in­creas­ingly be­ing chal­lenged in courts around the coun­try.”

The pro­cure­ment no­tice says the depart­ment is al­ready “con­duct­ing ex­per­i­ments in de­cep­tive be­hav­ior and col­lect­ing hu­man odor sam­ples.” The re­search it hopes to fund “will con­sist pri­mar­ily of the anal­y­sis and study of the hu­man odor sam­ples col­lected to de­ter­mine if a de­cep­tion in­di­ca­tor can be found.” stage of tech­no­log­i­cal de­vel­op­ment.

The direc­torate “is try­ing to de­ter­mine what fac­tors of hu­man be­hav­ior and chem­istry can pro­vide clues to the in­tent to de­ceive,” she said. The work would be car­ried out by the Fed­er­ally Funded Re­search and De­vel­op­ment Cen­ter run by the non­profit Mitre Corp., which con­ducts cut­tingedge re­search for the U.S. mil­i­tary, Home­land Se­cu­rity and in­tel­li­gence agen­cies, Ms. Kudwa said.

Sci­en­tific re­search shows that so-called volatile or­ganic com­pounds present in hu­man sweat, saliva and urine can be an­a­lyzed us­ing a tech­nique known as gas chro­matograph­mass spec­trom­e­try.

Re­search pub­lished by the Royal So­ci­ety in Lon­don in 2006 found “a sub­stan­tial num­ber of marker com­pounds [in hu­man sweat] that can po­ten­tially dif­fer- en­ti­ate in­di­vid­u­als or groups.”

Re­searchers took five sam­ples each from 179 peo­ple over a 10-week pe­riod and an­a­lyzed them, find­ing hun­dreds of chem­i­cal mark­ers that re­mained more or less con­stant for each per­son over time. An anal­y­sis of th­ese com­pounds “found strong ev­i­dence for in­di­vid­ual [odor] fin­ger­prints” the re­searchers con­cluded.

They warned, how­ever, that some peo­ple ap­pear to have less dis­tinc­tive odors than oth­ers.

“The rea­son for the vari­a­tion in dis­tinc­tive­ness is un­clear,” the re­searchers said. More im­por­tantly, the odors of some peo­ple changed dur­ing the course of the study. “Not all sub­jects had con­sis­tent marker com­pounds over time, which might be due to phys­i­o­log­i­cal, di­etary, or other changes,” they con­cluded.

The re­searchers also cau­tioned that some of th­ese marker com­pounds might be “ex­oge­nous chem­i­cal con­tam­i­nants,” from skin care or per­fume prod­ucts, or to­bacco smoke and other sub­stances present in a per­son’s en­vi­ron­ment. About a quar­ter of the 44 ap­par­ently dis­tinc­tive marker com­pounds they were able to an­a­lyze ap­peared to be ar­ti­fi­cial con­tam­i­nants, the re­searchers said.

“De­ter­min­ing the ori­gins of in­di­vid­ual and sex-spe­cific odors — and con­trol­ling ex­oge­nous chem­i­cal con­tam­i­nants — may pro­vide the most im­por­tant chal­lenge for fu­ture [. . . ] stud­ies,” the re­searchers said.

An­a­lysts said those chal­lenges are likely to be sig­nif­i­cant, and they will mul­ti­ply if the tech­niques are de­ployed in the field.

“While some of th­ese sen­sors per­form well in the lab, the real world may be dif­fer­ent,” said tech­nol­ogy con­sul­tant and au­thor John Vacca. “The tech­nol­ogy is still in its in­fancy.”

Newspapers in English

Newspapers from USA

© PressReader. All rights reserved.