A device promoted as a law enforcement tool to help bloodhounds detect human scents at crime scenes has come under increasing fire after its use in recent years has led to the incarceration of at least five men whose cases were later dismissed.
In the latest case, a Buena Park man was freed from prison in October after serving almost a year for a carjacking and armed robbery he did not commit. He was released only after a man jailed in Los Angeles County admitted to the crime, a confession supported by DNA evidence.
The scent-lifting device, known as a scent-transfer unit, or STU-100, was invented in the 1990s by Newport Beach engineer Larry Harris and a partner who has since died. Harris and a small band of Southern California supporters -- a group derided by the bloodhound-handling world as "social outcasts" -- have promoted the $900 machine to law enforcement.
Its backers say the machine, which resembles a leaf blower, can collect human scent from an object as small as a bullet fragment and transfer it to a 5-by-9-inch gauze pad that is put to a bloodhound's nose. The dog then theoretically follows the scent to the suspect.
The machine allows the scent to be presented to the dog without compromising physical evidence. But there is debate in the academic community over whether bloodhounds can reliably identify a specific suspect by his scent under any conditions.
It is unknown how many arrests or convictions can be attributed, at least in part, to the device because most law enforcement officials, including the FBI, declined to comment or did not return phone calls.
Civilian dog handler Ted Hamm, for one, says he has used the device in most of the 2,000 cases he has worked.
Despite its heavy use in law enforcement, critics say there is no hard evidence that the device works.
"I think it's quackery," said Larry Myers, professor at Auburn University's College of Veterinary Medicine and an expert defense witness in scent evidence cases. The dog handlers "have no idea how reliable the machine is."
In 2003, a California appellate court limited the use of scent evidence in state trials, ruling that the device and its operators have to meet standards that are "generally accepted" as reliable by the scientific community -- a benchmark that has not been achieved. Still, law enforcement continues to use the machine to identify suspects or gather enough probable cause for an arrest or search warrant.
"It's just one piece of the big investigation puzzle," said Lt. Larry Lincoln, a Los Angeles County sheriff's homicide investigator.
Agencies including Irvine, Buena Park and Long Beach police, the arson unit in Riverside County and the FBI have used the device.
I. Lehr Brisbin, a specialist in animal behavior and canine olfaction, said tests had shown that bloodhounds -- despite their uncanny ability to detect human smells -- are often unable to sniff out the originator of a scent among a group of people.
"The folks making these claims never allow their dogs to be tested" independently, said Brisbin, a professor emeritus at the University of Georgia. "But I don't think it's intentional fraud. It's a case of ultimate faith and belief in the machine and dog, but not in science. It's like a religion."
Harris declined to comment for this article, but Brisbin and Myers said the handlers who use the scent-transfer unit had been unable to explain how it works. This has hampered the admissibility of evidence gathered by the machine in state courts.
"I'm a dog guy, not a scientist," said Hamm, the only dog handler using the device who would be interviewed for this story. He works with the Los Angeles County Sheriff's Department at $125 per hour. "I can only say that I know it works through a lot of informal training. If people are willing to try and spend time with the dogs, they'll end up having a lot of success" with the machine.
Defense attorneys complain that the real value of the device is to give police probable cause to arrest a suspect when there is no physical evidence linking him to a crime.
"Larry Harris' dog was the excuse the cops needed to zero in on my client," said Santa Ana attorney Scott Borthwick, who represented James Ochoa, the man who was falsely accused of carjacking and robbery and released from prison in October. "Never mind that he was innocent. And that the dog walked past his house four times. And that it was the only house on the block surrounded by cops. Where else was the dog supposed to go?"
Though Harris has sold the machine nationwide for a decade, it is best known for its use in Southern California, where about 10 operators -- known in law enforcement circles as "Harris' disciples" -- use the scent-transfer unit in criminal investigations. The 10, who include civilians and police officers, make up the Southern California Bloodhound Handlers Coalition.
Their machine and dogs have led to false arrests in several high-profile cases.
* In 1996, Irvine resident Earl Rhoney became the first person in California convicted of murder based on evidence collected by Harris and his machine. On the day he was to be sentenced, then-Superior Court Judge Tony Rackauckas, now the Orange County district attorney, threw out his conviction on grounds that the device was unreliable.
Prosecutors chose to try Rhoney a second time but were forced to drop murder and burglary charges when another judge ruled that the scent machine evidence was tainted and could not be used. There was no physical evidence linking Rhoney, who has proclaimed his innocence, to the crime, which remains unsolved. He spent 3 1/2 years in Orange County Jail.
* Jeffrey Allen Grant, a Long Beach resident, was arrested in 1998, suspected of being a serial rapist. Police used Harris' machine to allegedly match Grant's scent to one lifted from crime scene evidence. DNA tests cleared him four months later, and he was awarded $1.7 million by a federal jury in 2000. The real rapist was sentenced to 1,030 years in prison in 2004.
Long Beach police officials declined to discuss the case. Grant could not be reached for comment.
* In 2003, Josh Connole was arrested as a suspect in a string of arsons and vandalism at four SUV dealerships. Knight, a bloodhound owned by Hamm, followed the scent from a gauze pad provided by FBI agents to Connole's Pomona home. Hamm said the FBI used its own machine to vacuum the scent. The same dog, then owned by Long Beach police, was used in the Grant case, Hamm said.
Connole, who came to the FBI's attention as a result of a tip, was arrested despite warnings from a federal prosecutor that agents did not have probable cause. He spent four days in jail and sued the FBI for civil rights violations. In 2005, he was paid $100,000 by the FBI and $20,000 by the city of West Covina for his mistaken arrest. A Caltech student was convicted in 2004 and sent to prison for the crimes.
Connole could not be reached for comment.
* Following the appellate court's ruling in 2003, murder charges were dropped against Jose Marin Flores, a Palmdale man accused of a fatal bar stabbing, because the evidence collected by the scent-transfer unit wasn't allowed into the trial.
* Last year, Ochoa was arrested after Harris said he used his machine to obtain the scent from a cap and shirt left in a stolen vehicle by a carjacker. A bloodhound allegedly followed the scent to Ochoa's house two blocks away. DNA tests completed before the trial excluded him as the person who wore the items, but he was charged anyway. He served 10 months in prison before the DNA was matched to a man in custody in Los Angeles for another carjacking.
Orange County district attorney's spokeswoman Susan Kang Schroeder said prosecutors would now use evidence gathered by the machine only "if it can be proved to our satisfaction that it's scientifically reliable."
Two national organizations of bloodhound handlers have refused to endorse the device because of its questionable reliability.
Hamm defended the machine. He said a scent vacuumed into a gauze pad can be frozen in a plastic bag for years.
"We train our dogs to pick up scent from small items, like bullet casings and fragments; items that have been touched by someone," Hamm said. "The [scent transfer unit] has proven to be a very reliable method of collecting human scent."