Researchers Create Drone Swarms That Can Detect Gas Leaks, Other Threats

 The research team used modified “CrazyFlie” drones, which are 12 cm in diameter weighed only 37.5 grams.

The research team used modified “CrazyFlie” drones, which are 12 cm in diameter weighed only 37.5 grams. TU Delft

A new algorithm called “Sniffy Bug” shows how tiny drones will do dangerous work even in areas where they can’t use GPS.

A new research paper documents the creation of the first autonomous small drone swarm that detect gas leaks as well as other possible chemical threats and can map rooms without the aid of GPS.

The research may be particularly relevant to the military, which is increasingly interested in small drones that perform well together with little human control and in tight places where GPS can’t reach—like underground—as well as drones that can be used in situations where chemical weapons or other hazards pose a threat to humans. 

So-called nano quadcopters, very small four-propellor drones, are easy for soldiers to carry and deploy, and their small size makes them difficult for enemies to target. But their size also limits their capabilities. They’re too small to carry the power necessary to run the same sorts of algorithms that allow self-driving cars or larger drones to map their environment, and they lack the energy to run things like light detection and ranging, which also helps autonomous vehicles to map their surroundings.

In a new paper published by IEEE, the researchers—from Netherlands, Spain, and Harvard University—discuss an algorithm that allows very small drones to map an area and even find the source of gas leaks. 

“Actually, in nature there are ample examples of successful navigation and odor source localization within strict resource constraints.” Bart Duisterhof of TU Delft said in a university announcement about the work. “Just think of how fruit flies with their tiny brains of ~100,000 neurons infallibly locate the bananas in your kitchen in the summer. They do this by elegantly combining simple behaviors such as flying upwind or orthogonally to the wind depending on whether they sense the odor. Although we could not directly copy these behaviors due to the absence of airflow sensors on our robots, we have instilled our robots with similarly simple behaviors to tackle the task.”

The algorithm, which they dub “Sniffy Bug,” works by sending the drones out to fly in “chainsaw” patterns to find the walls of the area in which they’ve been deployed. They can also detect one another in the air to avoid collision and collectively map the space. They can communicate with one another when one detects the source of a gas leak or other potential threat.

The algorithm could help first responders or soldiers working underground or in dense urban settings where it’s hard for drones to get GPS signals. The researchers describe it as “the first fully autonomous swarm of gas-seeking nano quadcopters.”