Search-and-Rescue Drone Locates Victims By Homing in on Their Phones

SARDO is a drone that flies over disaster zones and acts as a mobile cellular base

2 min read
screenshot from SARDO app
Image: SARDO

When a natural disaster strikes, first responders must move quickly to search for survivors. To support the search-and-rescue efforts, one group of innovators in Europe has succeeded in harnessing the power of drones, AI, and smartphones, all in one novel combination.

Their idea is to use a single drone as a moving cellular base station, which can do large sweeps over disaster areas and locate survivors using signals from their phones. AI helps the drone methodically survey the area and even estimate the trajectory of survivors who are moving.

The team built its platform, called Search-And-Rescue DrOne based solution (SARDO), using off-the-shelf hardware and tested it in field experiments and simulations. They describe the results in a study published 13 January in IEEE Transactions on Mobile Computing.

“We built SARDO to provide first responders with an all-in-one victims localization system capable of working in the aftermath of a disaster without existing network infrastructure support,” explains Antonio Albanese, a Research Associate at NEC Laboratories Europe GmbH, which is headquartered in Heidelberg, Germany.

The point is that a natural disaster may knock out cell towers along with other infrastructure. SARDO, which is quipped with a light-weight cellular base station, is a mobile solution that could be implemented regardless of what infrastructure remains after a natural disaster. 

To detect and map out the locations of victims, SARDO performs time-of-flight measurements (using the timing of signals emitted by the users’ phones to estimate distance). 

A machine learning algorithm is then applied to the time-of-flight measurements to calculate the positions of victims. The algorithm compensates for when signals are blocked by rubble.

If a victim is on the move in the wake of a disaster, a second machine learning algorithm, tasked with estimating the person’s trajectory based on their current movement, kicks in—potentially helping first responders locate the person sooner.   

After sweeping an area, the drone is programmed to automatically maneuver closer to the position of a suspected victim to retrieve more accurate distance measurements. If too many errors are interfering with the drone’s ability to locate victims, it’s programmed to enlarge the scanning area.

In their study, Albanese and his colleagues tested SARDO in several field experiments without rubble, and used simulations to test the approach in a scenario where rubble interfered with some signals. In the field experiments, the drone was able to pinpoint the location of missing people to within a few tens of meters, requiring approximately three minutes to locate each victim (within a field roughly 200 meters squared. As would be expected, SARDO was less accurate when rubble was present or when the drone was flying at higher speeds or altitudes.

Albanese notes that a limitation of SARDO–as is the case with all drone-based approaches–is the battery life of the drone. But, he says, the energy consumption of the NEC team’s design remains relatively low.

The group is consulting the laboratory’s business experts on the possibility of commercializing this tech.  Says Albanese: “There is interest, especially from the public safety divisions, but still no final decision has been taken.”

In the meantime, SARDO may undergo further advances. “We plan to extend SARDO to emergency indoor localization so [it is] capable of working in any emergency scenario where buildings might not be accessible [to human rescuers],” says Albanese.

This article appears in the April 2021 print issue as “Search-and-RescUAVs.”

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
LightGreen

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

"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.

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