A few weeks ago, the Innocence Project of New York (IP) announced that it had helped to release another innocent person from prison. This time it was Gerard Richardson. As The Verge outlined in September, Richardson was convicted of murdering a New Jersey woman in 1994 after a forensic odontologist concluded that the shape of Richardson’s jaw and the orientation of his teeth matched a bite mark on the murdered woman’s back.

After years of legal wrangling, IP was finally allowed to conduct a DNA test to double-check the odontologist’s conclusions. Their goal: to determine whether saliva swabbed from the bite mark in 1994 matched Richardson’s genetic makeup. It didn’t. The odontologist’s conclusion was proven false. Given this new information, prosecutors dropped Richardson’s case, and a judge declared him exonerated. After serving 19 years behind bars for a crime he had nothing to do with, Richardson finally walked home to his family as a free man on December 17th.

Richardson’s ordeal is but one in a steadily increasing number of cases overturned using DNA evidence. To this day, IP is aware of 311 such exonerations — cases in which someone was declared innocent of a crime long after they had been convicted in court. About 70 percent of those exonerations relied on DNA evidence.

But if you ask David A. Harris, that number should be much higher. Not only that; he also says new technology could accelerate such exonerations now — if only law enforcement would make the decision to use it.

DNA has exonerated hundreds of innocent prisoners

Harris is a law professor at the University of Pittsburgh who focuses on police behavior and law enforcement regulation. Last year, he published his third book, Failed Evidence, which argues that people in law enforcement are not only late to adopt state-of-the-art technology and scientific breakthroughs, they’re also fundamentally resistant to new innovations and to science.

An example of this, he writes, can be found in law enforcement’s approach to DNA.

While DNA evidence is often seen as a hyper-advanced, solve-anything, find-anyone technology from its portrayal on TV shows like Law and Order: SVU and Dexter, many of the FBI’s standards for DNA analysis are nearly 20 years old. (The FBI’s current standards cite a committee report from the National Research Council titled, An Update: The Evaluation of Forensic DNA Evidence. That update occurred in 1996.)

Those FBI standards, Harris writes, are way out of date. And as those standards sink further into obsolescence, many criminal cases are likely going unsolved, and many faulty convictions are likely going unchallenged.

Today’s forensic DNA analysis relies on "rules and procedures set up to allow relatively easy processing by lab personnel," Harris explains. Human beings, in other words, are expected to interpret DNA evidence one piece at a time, and then explain their analysis in court. Because current standards require a human being, and not a computer, to interact with every DNA sample, the analysis is labor intensive and can take a long time to carry out. As a result, crime labs can get bogged down if there’s a lot of evidence to analyze. And most DNA evidence gets tossed aside because it’s too complicated for a human to interpret.

"DNA evidence is much more complex than most labs can handle."

About that: an ideal DNA swab has one clear contributor. The Gerard Richardson case is a perfect example: one unknown person bit into a known person’s flesh. Analysts needed only to determine whether Richardson’s DNA matched that of the unknown person. It didn’t, and so his charges were dropped.

But when the number of unknown contributors to a DNA sample goes above one — or when the sample is tarnished in some way, or determined to be too miniscule to be analyzed by a person — there’s not much than can be done under current standards. Complicated DNA samples are thus called "uninterpretable" and often ignored.

This is no secret, either. Even the quintessential document about forensic science written in the last decade — the National Academy of Sciences’ (NAS) Strengthening Forensic Science in the United States, published in August, 2009 — laments that "DNA tests performed on a contaminated or otherwise compromised sample cannot be used reliably to identify or eliminate an individual as the perpetrator of a crime."

In reality, that’s not true anymore. DNA analysis has moved way beyond both FBI standards and NAS’ depiction of DNA’s limits.

Harris points to Cybergenetics, a company based down the road from his university office in Pittsburgh, that’s developed software called TrueAllele. An allele is a gene form that helps distinguish one person’s DNA as unique. TrueAllele uses algorithms rather than the naked eye to identify contributors to a DNA sample. Because it’s computer-based, it can interpret a lot of DNA evidence at once. The technique can also interpret DNA evidence currently considered tarnished. And it can identify as many as five unknown contributors to a DNA sample instead of just one.

"DNA evidence is much more complex than most labs can handle," says Dr. Ria David, one of Cybergenetics’ principals, who says as much as 80 percent of the evidence collected at crime scenes gets thrown away or cast aside. Cybergenetics’ software has been used in a number of high-profile cases, she says, but its use is limited.

According to Harris, law enforcement’s resistance to accepting computer-based DNA analysis amounts to a travesty. Think about it: if Dr. David is correct, and 80 percent of the DNA evidence collected at crime scenes today is neglected, what would happen if even half of that neglected evidence were able to be tested going forward? "You would get many more convictions and probably many more exonerations, too," Harris says. "If you thought DNA was a powerful tool for finding the truth, [Cybergenetics’] method is just that much more powerful and precise."

"If you thought DNA was a powerful tool for finding the truth..."

Harris sees computer-aided DNA analysis as "indisputably better" than what’s available today. (The technical descriptor for Cybergenetics’ software is "automated short tandem repeat STR analysis," and other companies such as the Center for Advanced DNA AnalysisBode Technology, and ZyGem are making similar strides forward.) "In five years," he predicts, "law enforcement will have no choice. But that moment hasn't come yet." He imagines it could arrive, however, as more prisoners such as Gerard Richardson force courts to allow the retesting of old DNA evidence.

"It’s my belief that we’re just at the tip of the iceberg when it comes to exonerations," he says. "Three hundred may not sound like a lot in comparison to the number of people who are in prison right now. But the number is only going to rise as advanced DNA analysis finds its way into the hands of more law enforcement agencies, more prosecutors, more defense attorneys."