LAPD’s fingerprint lab isn’t up to the task

Erroneous iden­ti­fi­ca­tions occurs some­times in foren­sic sci­ence.  Look at the Madrid bomber inci­dent in Spain.  I’m sure the LAPD will be tak­ing appro­pri­ate cor­rec­tive actions with the staff that was envolved in this inci­dent to ensure it doesn’t hap­pen again.

This more dis­turb­ing aspect of this arti­cle, is the poor chain of cus­tody pro­ce­dures described in the arti­cle.  Evidence should always be under lock-and-key, with lim­ited access.  Once the integrity of the evi­dence has been lost due to inse­cure chain of cus­tody pro­ce­dures, any results of analy­sis are questionable.

Original arti­cle posted here.

LAPD says staffers strug­gle to han­dle the work­load. Missing files and poor over­sight plague the cru­cial unit.
By Joel Rubin and Richard Winton
November 17, 2008
Late on the morn­ing of April 14, 2006, a trou­bling let­ter rolled off the fax machine in the har­ried, dis­or­dered fin­ger­print unit of the Los Angeles Police Department.

Months before, one of the unit’s print spe­cial­ists had deter­mined that sev­eral prints lifted from a cell­phone store where a bur­glary had occurred belonged to Maria Maldonado, a 25-year-old hos­pi­tal tech­ni­cian. Two oth­ers in the unit had signed off on the work. The match had given author­i­ties the evi­dence they needed to arrest the woman and charge her with the crime. When the case went before a judge, how­ever, a renowned fin­ger­print expert tes­ti­fied that the police had made a mistake.

The dis­trict attorney’s office sent the fax demand­ing answers.

The ana­lysts stood by their work, but days later the file con­tain­ing the sus­pected burglar’s prints mys­te­ri­ously dis­ap­peared from the unlocked drawer where it was kept. Working from copies of the prints, oth­ers in the unit and out­side con­sul­tants later con­cluded that Maldonado had, in fact, been wrongly accused, and the charges were dropped.

The case offers a stark pro­file of a high-stakes oper­a­tion that for years has been marred by inad­e­quate train­ing, anti­quated facil­i­ties, poor super­vi­sion, care­less han­dling of evi­dence and other short­falls, accord­ing to inter­nal police records and interviews.

“We were try­ing to hold on by the skin of our teeth to make sure things were done right,” said Diana Castro, the latent print unit super­vi­sor who retired early last year. “Were we per­fect? No. Did we do every­thing right? No. We did what we could to keep the cases flow­ing — to assist the detec­tives and the community.”

As focus in law enforce­ment has turned increas­ingly to the promise of DNA analy­sis to solve crimes, the fin­ger­print unit has lan­guished as a neglected but heav­ily used cor­ner­stone of the LAPD. It oper­ates 24 hours a day, seven days a week, with about 80 foren­sic print spe­cial­ists rotat­ing through three shifts.

They are sum­moned by detec­tives to roughly 24,000 crime scenes every year and about 60% of the time are able to col­lect fingerprints.

Prints are scanned into auto­mated data­bases that gen­er­ate dozens of pos­si­ble matches to known peo­ple. Specialists then seek a more pre­cise match by com­par­ing the prints under a mag­ni­fy­ing glass, look­ing at the dis­tinct ways in which the ridges of skin that make up a fin­ger­print split, stop and loop. Conclusive matches are made about 3,500 times a year.

The unit has long strug­gled to stay on top of its work­load, and work often backs up. Currently, fin­ger­prints from 3,018 prop­erty crime cases need to be entered into the com­puter data­base, and 1,052 cases are await­ing a more exact­ing man­ual analy­sis. The unit has not tended to requests from detec­tives try­ing to solve 320 homi­cides and other cold cases, accord­ing to depart­ment fig­ures pro­vided to The Times.

In the wake of the Maldonado case, senior LAPD offi­cials made some improve­ments to the unit but failed to address many other seri­ous shortcomings.

After The Times reported last month on an inter­nal LAPD report that high­lighted many of the ongo­ing prob­lems in the unit, offi­cials tight­ened the pro­ce­dure for ver­i­fy­ing fin­ger­print matches and, more recently, have reas­signed sev­eral print spe­cial­ists involved in misiden­ti­fy­ing Maldonado and a woman in another case, pend­ing an inves­ti­ga­tion of their work.

Even with these changes, how­ever, much still needs to be done to improve the effi­ciency and ensure the accu­racy of the unit’s work, said LAPD Deputy Chief Charlie Beck, who heads a recently formed task force inves­ti­gat­ing the fin­ger­print unit.

“It is a place that has been over­looked and pushed into a cor­ner,” he said. “That has to change.”

The unit’s facil­i­ties offer per­haps the most obvi­ous indi­ca­tion of the pri­or­ity the unit is afforded. As DNA work is con­ducted in a newly built lab­o­ra­tory, fin­ger­print com­par­isons are done in a con­verted kitchen on the eighth floor of the department’s down­town headquarters.

The unit’s auto­mated data­base work is done four flights below in a space shared with the unit’s admin­is­tra­tors. Fingerprint files are stored else­where. With too lit­tle room to have their own work space, three spe­cial­ists are assigned to each desk.

Beck, Castro and oth­ers described a chaotic, dis­rup­tive atmos­phere in which print spe­cial­ists must often break their con­cen­tra­tion to answer con­stantly ring­ing phones.

Long-standing plans to relo­cate the unit were scrapped in order to expand DNA lab facil­i­ties. The depart­ment and city agen­cies are search­ing for an alter­na­tive site for the print unit, Beck said.

That a file con­tain­ing orig­i­nal fin­ger­prints in the Maldonado case could van­ish illus­trates seri­ous flaws in how evi­dence has been han­dled for years. Specialists have been warned repeat­edly for years to be more dili­gent when sign­ing prints in and out of the unlocked fil­ing cab­i­nets where case evi­dence is stored, accord­ing to inter­nal records obtained by The Times.

“Court cases have been delayed because of this prob­lem,” a unit super­vi­sor told spe­cial­ists at a May 2006 meet­ing. Handwritten signs are com­monly posted on office walls when a file goes miss­ing, Castro and oth­ers with knowl­edge of the unit said.

In the past, mis­placed files were typ­i­cally found in trays on desks or in desk draw­ers. This year, spe­cial­ists were given secure lock­ers to store prints they analyze.

Three years ago, after being denied funds for a com­put­er­ized file-tracking sys­tem, a print spe­cial­ist with lit­tle com­puter knowl­edge was asked by super­vi­sors to build a rudi­men­tary one, Castro and Beck said. The in-house effort has been far from suc­cess­ful, how­ever, as spe­cial­ists reg­u­larly fail to update the sys­tem with infor­ma­tion about file where­abouts, they said.

“It is cer­tainly not a sys­tem for the 21^st cen­tury,” Beck said.

The Maldonado case is not the only instance in which prints were lost. In 1998, records show, detec­tives try­ing to solve a slay­ing case from a few years ear­lier asked the print unit to re-examine the fin­ger­print evi­dence. The unit could not find the prints and a fran­tic, weeks-long search proved unsuc­cess­ful. The prints have never been found, and the case remains unsolved, Beck said.

In a depart­ment in which audits and inter­nal inves­ti­ga­tions are used reg­u­larly to detect pos­si­ble prob­lems, over­sight of the fin­ger­print unit has been uneven. Command of the tech­nol­ogy lab, of which the fin­ger­print unit is a part, changed hands five times in the six years Castro led the unit, she said. And within the unit each super­vi­sor over­sees about 15 spe­cial­ists — far too many, Beck says.

With insuf­fi­cient super­vi­sion, train­ing and pro­mot­ing spe­cial­ists has been done in a largely ad-hoc man­ner, Beck, Castro and oth­ers said. In 2000, Beck said, the depart­ment did away with the exam it had used to test spe­cial­ists who sought pro­mo­tion from col­lect­ing prints in the field to ana­lyz­ing prints.

The deci­sion to drop the test was made, he said, shortly after the depart­ment added three dozen new mem­bers to the unit and “man­age­ment was over­whelmed with over­see­ing the increase in examiners.”

Now, spe­cial­ists auto­mat­i­cally become qual­i­fied to make fin­ger­print matches after spend­ing three years col­lect­ing prints at crime scenes. Before they are allowed to make matches they must com­plete class­room instruc­tion and receive men­tor­ing, but the unit does not have a for­mal man­ual that spells out train­ing require­ments, pro­to­cols or any­thing else, Beck said.

The department’s “Band-Aid” approach to prob­lems in the fin­ger­print unit and its fail­ure to pro­duce a com­pre­hen­sive man­ual is deeply trou­bling, said David Grieve, retired train­ing coör­di­na­tor of the Illinois State Police lab and a for­mer edi­tor of the Journal of Forensic Identification.

“If you don’t have it writ­ten down, peo­ple are going to stray, and things are going to become fragmented.”

Efforts to eval­u­ate spe­cial­ists’ work have also fallen short. The unit can afford to have an out­side test­ing agency admin­is­ter and score pro­fi­ciency exams for only five spe­cial­ists each year, Beck and Castro said. The rest of the unit is tested, but the scores are tab­u­lated inter­nally by a mea­ger staff of super­vi­sors and, until 2006, there were no con­se­quences for fail­ing. Between 2003 and 2006, Beck said, seven spe­cial­ists failed pro­fi­ciency exams. (Currently, a spe­cial­ist who fails the annual exam is taken off ana­lyt­i­cal case­work until he can be retrained.)

Beck said the depart­ment hoped to make suf­fi­cient improve­ments to the unit’s facil­i­ties and oper­a­tion in com­ing years to earn accred­i­ta­tion by the national board that gov­erns crime labs. He declined to com­ment on who in the depart­ment was to blame for the unit’s failures.

“For what­ever rea­son, this is some­thing that hasn’t received suf­fi­cient atten­tion,” he said.