Self-Learning Robots Produced With 3D Printers

Researchers at Oslo University have utilised 3D printers to make self-learning robots, capable of analysing a situation or problem, and developing a strategy to overcome the challenge.

The researchers envisage the use these types of robots in potentially hazardous environments, with limited human oversight.

Hazardous Duties

The Norwegian researchers have already developed the self-learning robots. According to Phys.org, the researchers believe that the robots could be used in hazardous environments, and cited examples such as working in a damaged nuclear power plant, landslide areas, or even mines on a faraway planet.

The robots have the capacity to overcome certain problems, such as moving past barriers and other obstacles.

“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,” Associate Professor Kyrre Glette was quoted as saying. Professor Glette is part of the Robotics and intelligent systems research team at Oslo University’s Department of Informatics.

“These environments are so extreme that no human being can cope,” he added. “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 researchers have reportedly already developed three generations of these self-learning robots.

The first generation, codenamed ‘Henriette’ was actually created ten years ago by Professor Mats Høvin. Henriette apparently had to teach itself how to walk and to jump over obstacles. If it lost a leg, it had to teach itself how to hop using just one leg.

The second generation robot came a couple of years later from Masters student Tønnes Nygaard, and came with a simulation program that allowed the robot to determine what its body should look like. The program allowed the robot to design how many legs it should have, the length of the legs and the distance between them. Essentially, the robot designed key parts of itself.

The third generation of self-learning robot incorporate much more flexibility, as its simulation program allowed for the robot to plot its complete design, including optimal number of legs and joints. This generation essentially came a self-learning, but more importantly, self-repairing robot.

Rise Of The Machine

“We tell the simulation program what we would like the robot to do, how fast it should walk, its size and energy consumption,” Professor Glette reportedly said. “For instance, we may want the robot to be able to turn around and change direction, climb over boulders and walk on rugged ground.”

Thus the simulation program analysed the requirements of each particular robot and its typical duties, and ran thousands of simulations to come up with the most optimal solution, which included the shape of the body and number of legs. The robots were all produced via 3D printing and tested by the researchers.

But the testing revealed that in reality, the robots had to cope with multiple issues, especially as their duties became ever more complex. But when the robots practised certain scenarios, they improved their performance by 20 to 40 percent.

The researchers are currently comparing the performance of five robots, each with different designs, over an obstacle course, to enable the robots to teach themselves how to pass the hurdles. The thinking is that in the future, the robots themselves could even carry a 3D printer, to allow the robot to ‘print’ the part they need to complete a particular task.

Earlier this week, a Deloitte report warned that over a third of the UK’s workforce or 11 million jobs could be at threat over the next few decades as computers and robots begin to take over many jobs.

But fear not….it is not all bad news. Researchers have developed a tiny robot that can swim through the bloodstream of a human being in order to deliver medicine to the problem area.

And last month Google revealed that it is developing a ‘nanoparticle’ pill which once swallowed, will enter the bloodstream and “sniff out” the dangerous chemical signals given off by cells as they become diseased.

Are you a Robot expert? Take our quiz!