These new self-learning robots could help explore alien worlds (or kill us all)

Aside from the whole artificial intelligence quotient, one of the hardest things about making useful robots is the basic utility of traversing the varied terrain on our planet. But these cutting-edge little critters take us one step closer.

Researchers at the University of Oslo have designed a new concept for robots capable of adapting to different environments, and the idea is a novel one. According to Popular Science, they’ve coded a computer program to design robots within specific parameters, such as the ability to move at a certain speed or climb over treacherous terrain.

But that’s just the beginning: After cranking out some designs, they go toe to toe in a series of virtual challenges to determine which of the designs would fare best in the real world. Once the algorithm has determined a winner, the 3D printer cranks it out and starts live testing. If something breaks during testing? It knows to bring a new part, or tweak the design to work anyway.

Here’s what Associate Professor Kyrre Glette, who is part of the Robotics and intelligent systems research team at Oslo University’s Department of Informatics, had to say about the project:

“In the future, robots must be able to solve tasks in deep mines on distant planets, in radioactive disaster areas, in hazardous landslip areas and on the sea bed beneath the Antarctic. These environments are so extreme that no human being can cope. Everything needs to be automatically controlled. Imagine that the robot is entering the wreckage of a nuclear power plant. It finds a staircase that no-one has thought of. The robot takes a picture. The picture is analysed. The arms of one of the robots is fitted with a printer. This produces a new robot, or a new part for the existing robot, which enables it to negotiate the stairs.”

The implications are awesome. You could send one of these to an alien planet to explore on the fly and send back tons of data no matter what terrain it encountered. Of course, a robotic system that can essentially design new robots and produce them itself sounds eerily like Skynet, or the evil robot baddies in The Matrix. So, yeah, c’mon guys, don’t give this thing too much autonomy.

(Via Popular Science, Apollon)

More from around the web