Drones as Detectives: Surveying Crime Scenes for Evidence

Researchers in Brazil are developing a drone that scouts for evidence—and want to use its footage to reconstruct crime scenes

2 min read
Still image from a video simulation of AirCSI over a crime scene.
Image: Pompilio Araujo

When detectives and other forensics specialists arrive at a crime scene, there is a pressing need to survey the area quickly. Environmental disturbances such as wind or an incoming tide could ruin valuable evidence, and even the investigators themselves are at risk of contaminating the crime scene. Could a fleet of evidence-surveying drones be of help?

Pompílio Araújo, a criminal expert for the Federal Police of Brazil, is responsible for recording crime scenes exactly as found. In his other role as a researcher at the Intelligent Vision Research Lab at Federal University of Bahia, he is trying to make his first job easier by developing drones that can—very quickly—home in on a piece of evidence and record it from multiple angles.

The drone system, dubbed AirCSI, starts by scanning a crime scene with one big sweep, using a stereo camera and a visual self-localization and mapping (SLAM) system to monitor the drone’s position.

The AirCSI drone system Photo: Pompilio Araujo

“Initially, the drone [flies] at a height that can take a broad view of the crime scene and detect some larger pieces of evidence,” explains Araújo, who published this preliminary stage of research in a previous study. The drone system was first trained to detect guns, but could be trained to identify other types of weapons or evidence such as blood stains.

In his latest work, described in IEEE Geoscience and Remote Sensing Letters, Araújo and his colleagues developed a secondary system for the drone, which includes a second camera, trained to image evidence from multiple angles. In this newer version, AirCSI calculates a circular area for each piece of evidence that’s initially identified, considering its relevance and size. The drone system then calculates a zigzag trajectory and makes a series of additional sweeps in order to collect more detailed data on each piece of evidence.

“As a result, AirCSI provides a sketch with the localization of the evidences, as well as a detailed crime scene imagery,” says Araújo. His team used simulation software to test this newer version of AirCSI, and found that using multiple angles to detect evidence is up to 18 percent more effective than using only one angle.

While the researchers have yet to test the new, multi-angle approach beyond simulations, they expect to try it out in a real environment by the end of this year or early next year.

Meanwhile, they are working to overcome a few of the system’s limitations. For example, AirCSI can currently operate only in open environments with no obstacles. “In the future, we plan to do anti-collision routines and training for more criminal evidences,” Araújo says.

He also plans on developing a way to completely reconstruct crime scenes using the drone footage, creating a virtual environment that investigators can explore indefinitely—or at least until the crime is solved.

The Conversation (0)

How the U.S. Army Is Turning Robots Into Team Players

Engineers battle the limits of deep learning for battlefield bots

11 min read
Robot with threads near a fallen branch

RoMan, the Army Research Laboratory's robotic manipulator, considers the best way to grasp and move a tree branch at the Adelphi Laboratory Center, in Maryland.

Evan Ackerman

“I should probably not be standing this close," I think to myself, as the robot slowly approaches a large tree branch on the floor in front of me. It's not the size of the branch that makes me nervous—it's that the robot is operating autonomously, and that while I know what it's supposed to do, I'm not entirely sure what it will do. If everything works the way the roboticists at the U.S. Army Research Laboratory (ARL) in Adelphi, Md., expect, the robot will identify the branch, grasp it, and drag it out of the way. These folks know what they're doing, but I've spent enough time around robots that I take a small step backwards anyway.

This article is part of our special report on AI, “The Great AI Reckoning.”

The robot, named RoMan, for Robotic Manipulator, is about the size of a large lawn mower, with a tracked base that helps it handle most kinds of terrain. At the front, it has a squat torso equipped with cameras and depth sensors, as well as a pair of arms that were harvested from a prototype disaster-response robot originally developed at NASA's Jet Propulsion Laboratory for a DARPA robotics competition. RoMan's job today is roadway clearing, a multistep task that ARL wants the robot to complete as autonomously as possible. Instead of instructing the robot to grasp specific objects in specific ways and move them to specific places, the operators tell RoMan to "go clear a path." It's then up to the robot to make all the decisions necessary to achieve that objective.

Keep Reading ↓ Show less