The deputy director of the Kansas Bureau of Investigation says critics have knocked the bureau's new $2 million fingerprint computer because only 17 suspects have been identified by the system since it went on line in September.
But the deputy director, Steven Starr, and a local detective who has used the system say they're impressed with the computer's performance so far and think it will be a boon to law enforcement in the future.
John Lewis, the first Lawrence police detective to use the system, said the computer helped an investigation into a string of armed robberies in Lawrence and other Kansas cities even though it didn't identify the man who eventually was arrested for the crimes.
Lewis said the computer showed local officers that a man they were investigating was not the suspect. That saved days of investigation and allowed officers to shift their focus to another suspect.
"IF WE HADN'T had the system," Lewis said, "we would have had to send in the prints and they (KBI) would have had to compare them manually. And we would still have had to focus our attention on somebody who didn't have anything to do with the robberies."
The computer stores both "ten prints," which are cards turned in by jailers and other arresting officers, and "latent" prints, impressions taken off of evidence such as weapons, doorknobs and recovered stolen items.
When someone is arrested and fingerprinted, the law enforcement agency that made the arrest sends the prints to the KBI. Specialists then enter an image of the prints into the computer, which scans the prints and compares them with similar ones in the file of "latent" fingerprints. If the computer finds a match, then the person who has been arrested may become a suspect in the crime that generated the latent print.
If law enforcement officers turn in a latent fingerprint, the computer can search the "ten print" files to see if they contain the identity of the person who left the fingerprint.
KBI OFFICIALS will not use the computer alone to determine if prints match. If the computer identifies a match, KBI officials retrieve the actual fingerprint from their files and make a manual comparison.
So far, the computer has brought breaks in two major cases.
Starr said the computer, known as the Automated Fingerprint Identification System, or AFIS, identified a suspect in a Sedgwick County sheriff's department investigation into a string of rapes.
The computer also identified a suspect in the murder of a Wamego school teacher.
Before the system was installed, comparing existing fingerprints and latent fingerprints was a laborious, time-consuming process.
When a local law enforcement agency wanted to compare a latent print to a ten print, the quickest way was to identify a possible suspect. KBI officials then would check their files, which contain 350,000 ten print cards, for the suspect's name.
IF THE KBI had the suspect's ten prints, officials would pull the card and compare the prints on it to the latent print.
When a possible suspect's name wasn't available, law enforcement agencies could use a KBI categorization system that narrowed the possible list of matching prints. For instance, there are four categories of fingerprints, defined by the print's basic pattern. In each pattern, there are further similar characteristics.
But even when incoming prints were categorized, KBI experts had to manually search thousands of cards.
Starr said AFIS can search the system in a fraction of the time a manual search would take. Among the system's other features, it can lighten or darken the image of a print, shade certain areas of the image and enlarge or shrink the image.
What the system can't do, Starr said, is solve any case in which detectives find a fingerprint.
"FOR STARTERS, not everyone in the United States or even in the state of Kansas has been fingerprinted," he said. Therefore, the computer often won't have prints to compare to a fingerprint submitted by law officers.
Currently, not all Kansans who have been fingerprinted are in the KBI's AFIS. Workers are trying to enter a backlog of more than 20,000 cards into the system.
Once the cards are entered and AFIS is on line, Starr and Lewis say there will be a better chance that mysterious crimes will be solved.
"I can foresee us using it on any major case," Lewis said. "I can foresee that being one of the first things we do. It's not going to solve all our cases, but it's going to save time in a lot of them."