Using AI, a robotic hummingbird can go where other drones can’t

Birds have always been inspirational drones for many years. While one of the natures’ most incredible creations Hummingbirds has something special. They can flap their wings so fast, about 80 times per second and can fly right, left, up, down, backward, and even upside down. Its ability to hover and make sharp turns could teach drones a few things.

Using machine learning, the researchers at Purdue University studied for a long how these birds fly in order to replicate their abilities in drones.

And then the team has built a hummingbird inspired robot that behaves like hummingbirds, trained by machine learning algorithms based on various techniques the bird uses naturally every day. After learning from a simulation, the robot “knows” how to move around on its own like a hummingbird would, such as discerning when to perform an escape maneuver.

Drones cannot be shrunk to very small sizes, as they wouldn’t be able to generate enough lift to support their weight. But hummingbirds don’t use conventional aerodynamics – and their wings are resilient. And the drone made by the researcher team is nearly close to the size of the bird.

This robotic hummingbird flies on its own while tethered to an energy source, but will soon be powered on batteries. (Purdue University video/Bio-Robotics Lab)
This robotic hummingbird flies on its own while tethered to an energy source, but will soon be powered on batteries. (Purdue University video/Bio-Robotics Lab)

Hummingbird Robot Features

The hummingbird robot is made of carbon fiber and membranes which are cut with lasers, and its body is 3D-printed. It weighs only 12 grams (0.4 ounces) and also can lift more than twice of its own weight. The team has also created an insect-sized robot which weighed just 1 gram.

It can fly like a hummingbird and also can pull off some of the most insane aerial stunts like hovering and turning 180 degrees in 0.2 seconds flat.

These tiny drone doesn’t have any camera to see the surrounding, but still, through an electrical sense of touch and AI algorithms, it can track its movements around objects. It senses its environments by touching surfaces and learn how to move efficiently.

Xinyan Deng and her team of researchers (Purdue University photo/Jared Pike)
Xinyan Deng and her team of researchers (Purdue University photo/Jared Pike)

While these drones will not be the fastest and farthest fliers, their improved maneuverability and small size make it move through spaces that other robots can’t fit into. And, due to its small size and quiet operation, it can be well suited secret use as well as search and rescue operation.

The robot can essentially create a map without seeing its surroundings,” Xinyan Deng, an associate professor of mechanical engineering at Purdue University, explained in a statement. “This could be helpful in a situation when the robot might be searching for victims in a dark place — and it means one less sensor to add when we do give the robot the ability to see.

Before more sensors can be added to the robot, however, the team says that its lifting power needs a boost. Cameras, GPS and other sensors could be useful in future versions, and batteries would allow it to fly free – in its current form, the flying robot has to remain tethered to an energy source.

Trending

Smart helmet to help search-and-rescue teams save lives faster

Hypergiant's Futuristic Helmet keeps the military, police, firefighters and other first responders safer than ever.

FarmWise’ weed-pulling robots take care of each plant individually

FarmWise's robots use computer vision and machine learning to provide personalized care to every single plant they touch.

Related Stories

LEAVE A REPLY

Please enter your comment!
Please enter your name here

Stay on top - Get the daily news in your inbox

Get the best futuristic stories staight into your inbox before everyone else!