I see crime scenes
Fighting crime with science isn't as simple as popular TV shows like CSI would have you believe. But those shows get one thing right: science and technology have a tremendous potential for changing the way crime is investigated.
Today, investigators record crime scenes using sketches and photographs. Bullet trajectories are determined using lasers or even lengths of string. "They take a microsnapshot," said Mitzi Montoya, a North Carolina State University researcher who specializes in knowledge and virtual team management. "You don't really know what's relevant in a crime scene, and you can't go back and create it, because once it's cleaned it's gone forever."
That could change if the IC-CRIME (Interdisciplinary Cyber-enabled Crime Reconstruction through Innovative Methodology and Engagement) project is a success. IC-CRIME will use a crime scene capture device that records measurements of an entire scene using lasers coupled with high-resolution digital photography.
Although these scanners already exist on the market, IC-CRIME will take them further, uploading the scanner's data into a 3D game engine complete with avatars, simulation tools, and tools to enable collaboration between investigators everywhere.
Investigators will be able to use IC-CRIME to recreate crime scenes, analyze trajectories and blood spatter patterns, and simulate scenarios.
The project grew out of a friendly chat between Montoya and her co-principal investigator, David Hinks, an NCSU researcher and forensics expert who regularly consults for the North Carolina Federal Bureau of Investigations.
"As we talked about the challenges that face forensics, I realized that they're a classic virtual organization without an organization," said Montoya. In this case, the virtual organization is the team of forensic experts.
The third co-principal investigator for the project, Michael Young - also a friend of Montoya's - specializes in digital gaming research. "The lightbulb went off when I saw how our three areas of expertise could be connected in an innovative way to solve a problem," said Montoya.
The IC-CRIME team is also working on a comprehensive fiber database. On TV, you get the impression that investigators are always finding DNA, said Montoya. In reality, "latent evidence tends to be 70- to 80- percent of what you find on a crime scene, with fiber making up a large part of that." Latent evidence is any evidence that is present at the crime scene, yet cannot be seen with the naked eye.
At the moment, the best forensic scientists can do is say that two fiber samples are consistent with each other. But with a comprehensive database of fibers, investigators could really put that large percentage of evidence to good use.
"There are fundamental research questions related to this project," said Montoya. For example, what resolution of data should the system collect? What information should it keep, and what could be discarded? The scan of the scene would be most accurate if it is the first thing investigators do at the scene, before they start to disturb the scene searching for evidence. Yet the scan takes up to two hours, and there are other forms of evidence which degrade over time. So, what is the best way to preserve as much of the most crucial data as possible?
The three-year project, which received its funding on 15 September 2009, is using mock crime scenes for preliminary data while IC-CRIME is under development. The system will undergo more rigorous testing at North Carolina State University's annual law enforcement training program; part of the program is the enactment of a number of complex mock crime scenes, perfect for testing the IC-CRIME system.
If local law enforcement agencies are pleased with what they see at that point, the system will be tested with non-violent crimes before it finally graduates to its intended area of focus: homicide.
"A next step could be a cloud computing solution," said Montoya. She envisions a nationwide network of locally-maintained IC-CRIME databases which could be accessed from a single cloud-based interface. To provide multiple users with real-time access to all crime scene simulations hosted on a given database, each database would require its own cluster of computers.
IC-CRIME isn't just for investigators who wish to study a crime scene and assemble possible scenarios. Lawyers may wish to use it to argue their cases during a trial. Trainers and educators may choose to use it as part of their regular curriculum.
"Since there is no database of crime scenes complete with high-resolution geospatial data, no one has ever done pattern analysis on crime scenes," said Montoya. "Now we can."
-Miriam Boon, iSGTW