NIST Corner: Forensic Ballistics
Written by Robert M. Thompson   
Combing through a crime scene, an investigator finds fired cartridge cases. The crime lab later matches them to a handgun purchased locally by a likely suspect. At trial, the defense attorney casts doubt on the ballistics match, noting that the gun store sold five other handguns of the same model and the same manufacturing lot to other area residents. In this scenario, is there a way to determine whether or even how likely it is that the cartridge cases were actually fired from one of the other handguns?
A team of NIST researchers is working on this measurement challenge. Their ultimate goal is the ability to measure the degree of similarity between patterns left by consecutively manufactured firearms as well as degree of variability between pattern marks produced by the same weapon.
Achieving this level of measurement certainty about the microscopic identifying marks left on fired bullets and cartridge cases is the holy grail for firearms examiners. With it, firearms examiners would be able to quantify a level of confidence in their analyses of ballistics matches. As a result, juries would have additional objective information with which to weigh ballistic analyses, even against the possibility that a gun from the same manufacturing lot might have been used.
Xiaoyu Alan Zheng, a mechanical engineer in the NIST Semiconductor and Dimensional Metrology Division, along with the Forensic Toolmark Analysis Project team, is leading this effort to measure microscopic marks on bullets and cartridge cases at micrometer (millionth of a meter) scales. The team is using some of the same measurement techniques that the semiconductor industry uses to design and build the ultra-small processors in mobile phones, tablets, and laptops.
Their work expands on decades of research by forensic practitioners and academic researchers who have examined consecutively manufactured tools such as barrels, slides, chisels, punches, and knives, to name just a few.
“We certainly can apply these sophisticated measurement techniques to the challenge of improving the science behind ballistics matching,” Zheng says. “One of the critical pieces is new algorithms that can make sense of topographical measurements made at the micrometer level. The other is a reference set of topographical data from which academic and industry researchers can develop the improved algorithms.”
Historically, firearms examiners performed labor-intensive visual comparisons of patterns on bullets and cartridge cases from a crime scene with ballistics signatures stored in separate police department files. In the 1990s, automated ballistic identification systems emerged with support from the U.S. Bureau of Alcohol, Tobacco, and Firearms and the Federal Bureau of Investigation.
These systems allowed law enforcement agencies to digitize images of cartridges and bullets. They also automated the process of comparing marks on a bullet or cartridge case from a crime scene to a shared database of collected images. At the heart of these automated systems are mathematical algorithms that help sort and point to matching patterns.
One key to developing improved algorithms that can discern the differences between marks from consecutively manufactured weapons is accurate reference data. Researchers can use such data to test the performance of algorithms. With assistance from law enforcement agencies, the firearm examiner community, and manufacturers, NIST researchers are developing an open-access research database of precisely measured 3D topographic data from fired cartridge cases and bullets.
The NIST team has borrowed sets of consecutively manufactured weapons from law enforcement agencies and manufacturers and fired them at local crime laboratory firing ranges. The researchers then scan the surface of the fired bullets and cases using a disc scanning confocal microscope to collect the 3D surface topography data.
Another aspect of their research is evaluating whether an objective mathematical algorithm can accurately identify the tool that created a striated or impressed toolmark from a pool of consecutively manufactured tools. Last spring they published a study demonstrating that a mathematical metric—known as the Cross Correlation Function—correctly matched consecutively manufactured chisels and drift punches with toolmarks produced by each tool.
The study, published in the AFTE Journal (Zheng et al, Volume 46, Number 2, Spring 2014), evaluated whether the Cross Correlation Function could accurately identify unknown toolmarks back to the tool source.
Using a precise toolmarking robot, NIST researchers made marks onto copper plates with 20 chisels and 20 punches – two per chisel and two per punch. They knew the identity of the tool for 40 of the marks (one known mark for each chisel and punch). Then they randomized and hid the identities of the chisels and punches and made another set of unknown marks with each tool. They found that the algorithm accurately paired each unknown mark with the chisel or punch that created it.
“The results add support to the forensic science specialty of toolmark identification. With further research, these methods may one day contribute to a toolmark examiner’s testimony in court,” they wrote in the AFTE Journal paper.
For More Information

About the Author
This e-mail address is being protected from spam bots, you need JavaScript enabled to view it is the Senior Forensic Science Research Manager for NIST.
< Prev   Next >

Interview with an Expert

One of the more specialized areas of crime-scene investigation has to do with searching for evidence of arson. To get some background in this area, we spoke with an individual who has had more than 46 years in fire service, 24 of which have focused specifically on fire/arson investigation.