It is likely that very soon, robots will live in homes with people, helping older people to live independently. But for this they will have to learn to do all the little jobs that people could do without thinking. Many modern artificial intelligence systems are trained to perform certain tasks, analyzing thousands of signed images action. Although these methods help to solve increasingly complex problems, they still relate only to very specific needs and require a lot of time and computing power for training.
If the robot will take care of the elderly, the problems of this work will vary greatly when compared with typical situations in the learning process. During the day the robots have to do a lot of things, from tea to linen change during the conversation. These are complex challenges that become harder in the combination. No two homes are the same, which means the robots have to quickly learn and adapt to the environment. And as it often happens, if you live with someone else, things tend to migrate. The robot will have to learn to find them yourself.
One approach is to develop a robot able to learn throughout life, which could store knowledge based on experience, and develop ways of their adaptation and application to new challenges. Once you learn to make a Cup of tea, these skills can be applied to coffee.
The human brain learns throughout his life, constantly adapting to complex and changing conditions and solving a wide range of problems. Modeling how people learn, could help in the development of robots with whom we can interact in a natural way, like with another person.
Modeling of child development for learning robot
The first question to ask when starting to model people: where to start? Alan Turing, famous mathematician and pioneer of artificial intelligence, once said:
“Instead of trying to create a program to simulate the adult consciousness, why not try to produce software for the simulation of the child? If it were so, then after an appropriate course of education one would obtain the adult brain”.
He compared the brain of a child with a blank notebook, which can be filled in the education process and to develop intelligent adult “system”. But what should be the age of the child for the simulation? What knowledge and skills you need to lay first?
Newborn infants are very limited in what we can do and how to perceive the world around them. Muscle strength in the baby’s neck is not enough to support the head, and he has not learned to control their arms and legs.
Start with zero month — such a step may severely limit the robot. But the physical limitations of the child really help him focus on solving a small subclass of problems, for example, he learns to relate their eye that hears and sees. These steps at the initial stages of model building child build his body before he will begin to understand the complexity of the world around.
Engineers have applied similar restrictions to the robot, initially blocking movement of various joints, to simulate the lack of muscle control. They also adjusted the camera image of the robot to “see” the world through the eyes of a newborn — with a blur and a weak periphery. Instead of telling the robot how to move, let him figure it out on their own. The advantage here is that the changes of calibration, or as damage to the limbs, the robot is able to adapt to these changes and continue to work.
Learning by playing
Studies have shown that due to limitations in the learning process, increases not only the speed with which the acquired new knowledge and skills, but also increases the accuracy of what is being studied.
Providing the robot control deregulation — giving him control over the joints and improving its vision is to ensure that the robot will control its learning rate. Scientists have modeled “child” and the first 10 months of its growth. As soon as the robot was trained to correlate the movements and the resulting sensory information, he took on the stereotypical behavior that has been observed in infants — such as when children spend long periods of time, staring at their hands while driving.
When the robot learns to coordinate his own body, the next major milestone, which he is — he begins to understand the world around him. The game is an important part of a child’s learning. She helps him to explore the environment, test the various possibilities and examine the results.
At first it might be something simple like knock a spoon on the table or put in the mouth of any subject. But then it develops into building towers out of blocks or putting objects into the appropriate holes. All of these actions create experience, which will further provide the basis for skills such as finding the right key to open a lock and fine motor skills to insert the key into the keyhole, and then rotate it.
In the future the use of these methods will give robots the tools for learning and adaptation to difficult conditions and challenges that people take for granted in everyday life. One day robots will be able to assist the elderly, but to teach them even the children in kindergarten.
Don’t forget to discuss the robot in our chat in Telegram.