The Center for Statistics and Applications in Forensic Evidence (CSAFE) is creating innovative solutions to the lack of scientific foundations in firearms examination. Firearms examiners are charged with providing accurate assessments of bullet evidence, despite limitations in current analysis techniques. Current best practice relies on a subjective visual inspection of bullets and cartridge cases to determine if a particular gun was used at a crime scene. Due to potential human error and bias, this methodology decreases confidence in results when used in courtroom testimony.
CSAFE researcher Dr. Heike Hofmann, Professor of Statistics at Iowa State University, and contributing researcher Dr. Eric Hare have developed a groundbreaking new method of bullet comparison. The CSAFE researchers’ new bullet matching technology uses a fully automated statistical algorithm model to predict the probability of a match of two unknown pairs of bullets.
In response to the need for continuous improvement in the process of bullet comparison, Hofmann said, “We want to get away from language of ‘this is a match’ and ‘this is not a match’ and move towards a matching score that allows us to assess the quality of evidence on a continuum.”
Hofmann acknowledges that in some cases, matches are very convincing where everyone who examines the bullets agrees. “Then there’s those cases in-between where different forensic examiners might come to different conclusions,” she said.
Hare and Hofmann worked in collaboration with the National Institute of Standards and technology (NIST) to develop the algorithm utilizing the NIST ballistic toolmark database of 3D bullet scans. Researchers initially identify signature features or a combination of features on each bullet that carry the most information. The algorithm then numerically computes differences, enabling discrimination across samples.
Hofmann said, “The idea is that when you’ve got an automatic approach that you get a number at the end and that process is repeatable. So if you get the same input, use the same parameters, then you will always get the same results.”
Hofmann’s goal is to provide a quantitative assessment of the association of two bullets through a signature-based score, thus reducing human error. “You’re taking the human being a little bit out of the loop, and the matching becomes less subjective and moves into a more objective realm,” she said.
Public Access Increases Transparency
CSAFE researchers are breaking with tradition by publishing technology that is accessible to everyone through an open model. Hofmann and Hare’s algorithm can be found in their R package “bulletr”, which is publicly available on GitHub.
Hofmann said, “I totally believe in openness and transparency. I want to have transparent algorithms that can be peer reviewed and checked. I want to have open data so other people can have access to the same kind of information.”
At present, a suspect accused of a crime has no opportunity to review the results of firearm analysis beyond the 2D images of the crime scene and questioned samples. CSAFE’s new technology is user friendly enough that experts on both sides in a trial can carry out their own analysis of the data.
“We are implementing new features so they are accessible to everyone, not just a selected group,” Hofmann said.
This technology can also be used to assist juries in determining the strength of a match between bullets. With an automated algorithm model, examiners are able express a degree of uncertainty in their conclusions.
If the score from the algorithm is not as high as is seen in other known matches, the examiner has a way of saying that he or she feels that it could still be a match, but that the analysis is not 100 percent conclusive. A jury is able to see the score and determine if there is a measure of doubt in the results, thus protecting against wrongful convictions.
Future Collaborations Aim to Engage Practitioners in the Process
CSAFE researchers have already developed many collaborating partnerships in an effort to expand the reach of their technology and include practitioners in the conversation.
In order to continue creating objective and repeatable methods of bullet comparison, Hofmann and her team are designing a large-scale experiment in collaboration with the Story County Sheriff’s Office and the Defense Forensic Science Center.
The goal of the experiment is to continue expanding the database of processed bullets, and advance the development of the open source R package “bulletr.”
With a team based approach to research and commitment to learning from practitioners, CSAFE will continue to advance firearm examination technology with real-world applications.