Guessing Robots Predict Their Environments

Engineers at Purdue University developing countries are in a position of the robot “guess” about what we expected, since it through unknown, reduces the amount of time that he needs in order to successfully navigate in these environments.

The method works with the help of an algorithm, new software, with the assistance of a robot partly cards in circulation, as it is in a for the first time. The robot will be on this card partially predict what awaits us.

The repeated exposure to the environment, the accurate prediction and it is easy to navigate, in which robots with success, said CS George Lee, a Purdue professor of Electrical Engineering and Computer Science, which specializes in robotics.

“This is the navigation easier, a garage with this card, because every floor is identical or very similar, and the same could be some so-called office building,” he said.

Both simulated and robotic effectively used in the search for information about a laser range finder to measure and distance sensor for the environment and the creation of tickets for the drawing.

The algorithm is an approach called SLAM, which emerged in the years 1980. The name SLAM, for the simultaneous localization and mapping, was invented in 1990 by Hugh F. Durrant -Whyte and John J. Leonard, and the engineers at the University of Oxford, United Kingdom.

SLAM used data from sensors for the orientation of a robot, with the maps of the immediate surroundings. Since the new method used to predict the cards, which we expected, it is P-SLAM.

“The effectiveness depends on the presence of features repeated, and similar types of symmetric structures, such as walls rights, right-angle bends, and a presentation by the same premises,” said Lee. “This technique allows a robot, an assessment of the future, the proportion of the environment already drawn.”

The research findings were detailed in a document published in April in IEEE Transactions in the robotics, published by the Institute of Electrical and Electronics Engineers. This document was signed by the PhDs Jacky H. Chang, Lee, Professor Yung – Hsiang Lu and Professor Y. Charlie Hu, while in Purdue’s School of Electrical Engineering and Computer Science.

Potential applications include robotics, and in particular the military and fighting robot, which according to the buildings and other environments.

The Purdue researchers tested their algorithm in both robots and simulated robot in a real sailing the course in a building on the campus of Purdue. Conclusions have shown that the simulation of a robot with the algorithms could navigate the labyrinth, practically everything to explore 33 in percent less environment, which normally should be required.

The future expansion of the concept, four robots, a team, working like ants to an unknown environment, the information about the progress of wireless network. The researchers are also working on the creation of an object on the forecast, “admits things like doors and chairs, as well as to strengthen the robot” energy efficiency.

Robots work without the knowledge in the cards must be fully on the sensors for the leadership on the environment. These sensors are, but sometimes inaccurate, and mechanical errors of the robot durchstreifend something.

The algorithm of robots can to correct this error, which is on the map, the navigation with greater precision and efficiency.

“If the robot is a successively a corner, we say there is a degree of mechanical error, and it runs a little too strong or not strong enough,” said Lee. “Then, when the robot to travel further to the right, that small error is a very big mistake navigation system in the future.”

The search was launched by the National Science Foundation.

In the works, Purdue students in a conception of the class have a prototype of fire-fighting robot called Firebot.


One Response to Guessing Robots Predict Their Environments

  1. Sunehri says:

    Honda Robots are coming that will be controlled by Human Operators.
    They can also be acclimitized to control envoirments.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: