What happens when AI meets robotics?

Peter Stone

Researchers in Texas are developing robots that have a mind of their own.

The scientists build systems that can learn and interact in the family, workplace, and even on the sports field.

The University of Texas at Austin team is the inclusion of artificial intelligence into the machines, so that you can deal with real-life situations.

Among the systems, automated assistant, the implementation of simple tasks in an office work

Science-fiction movies predicted that we would have in the future, intelligent robots.

In the day the earth stood Still, we had the dark Gort; in Forbidden Planet, Robby was; and in the TV series Lost In Space Zachary Smith nemesis, the robot was.

Peter Stone

There are more than 50 years, since the fictional representations are so where are they?

Although we have robots in factories, the exit for decades, to the shop floor-not an easy task. In the production, they lead to predetermined, repetitive tasks, all day and night.

But not to cope when you step outside, you will be able to the chaos, in the real world. It is a place where order and routine are over. Even the simplest tasks are made difficult by the unpredictability and uncertainty of human interaction.

A team from the University of Texas, Austin has given itself the challenge of bringing robots out of their ” comfort zone and in our chaotic world

Laptops on motorized pedestals Hiking spooky about the lab as slowly, partially built, Cybermen or Daleks. The researchers are Wide, the building Intelligence project. Currently, you have to learn your way around the place.

Pallab Ghosh/ BBC News

So far you have learned how to avoid them bumping into things. But great things are planned for you.

The goal is for you simple instructions can perform a search for a colleague, or locate and bring over a piece of equipment. But this is easier said than done.

You must be able to recognize objects and people. You also need to understand, to recognize, instead of human speech. There is a difference. Later, a limited number of voice commands, sets pre-programmed. But understanding requires context and meaning.

This can only be achieved with AI techniques, in which the robot questions, to understand to help him, what words mean. The researchers leading the project is a linguist, Dr. Jesse Thomason. He has limited the number of questions, the partially built Daleks only ask five. Why only so few?

He tells me that the robot be “annoying” with their questions. They would ask you, for always like a little child.Top team

In fact, when Jesse teaches one of the robots, it seems, he speaks of a baby.

His colleague, Dr. Andrea Thomaz is trying to teach the human etiquette, the Daleks.

She wants you to know when it is appropriate, the proximity of people and start a conversation.

Dr. Thomaz has been designed to enable code to give out the robots, social cues, such as waving to someone, and then look for movements or expressions, to speak of their human counterparts, indicating that they are happy.

“The problem really is that robots will be able, with the dynamics and the noise and unpredictability that the people in the area, and so we have to think about the perception, control and learning, robot to build that can deal with this kind of thing,” she told BBC News.

In another part of the laboratory, in the humanoid robot to play soccer. You are part of an international “RoboCup” initiative, the challenge to build a team, able to beat the men-world Cup winners by 2050.

The robots play on a miniature indoor football field, work together as a team and learn from each game.Tiki-taka-robot

Playing football is much more difficult for a machine-learning systems who’s in charge of as chess or go, according to Prof Peter stone of this project and regularly plays football.

“Rather than turn-taking, each movement in the same time. And if you take too long to decide what to do, if you pass the ball, your opponents come in and take the ball.

“It is also continuous. There are no discrete places where people need to be. You are always in motion, through the air and through space.

“There are so many challenges unlike some of the Board games.”

The types of robots in the 1950s, the movies are still in the realms of science fiction. But with giant strides in the artificial intelligence, it will not take long and you will life to be a part of our daily.

Follow Pallab on Twitter