Information Systems

Virtual dog to teach robots to perform tasks intelligently

Washington, May 17 (IANS) While playing with your pooch at home helps evoke love and compassion in you, researchers from Washington State University are using a virtual dog to train robots how to do desired tasks.

The team has designed a computer programme that lets humans teach a virtual robot that looks like a computerised pooch.

Non-computer programmers worked with and trained the robot in the intelligent robot learning laboratory at the university.

"At the beginning, the virtual dog moves slowly. But as it receives more feedback and becomes more confident in what to do, it speeds up," said Bei Peng, doctoral student in computer science.

For the study, the researchers varied the speed at which their virtual dog reacted.

"As when somebody is teaching a new skill to a real animal, the slower movements let the user know that the virtual dog was unsure of how to behave. The user could then provide clearer guidance to help the robot learn better," the authors noted.

The user taught tasks by either reinforcing good behaviour or punishing incorrect behaviour.

The more feedback the virtual dog received from the human, the more adept the robot became at predicting the correct course of action.

The researchers' algorithm allowed the virtual dog to understand the tricky meanings behind a lack of feedback - called implicit feedback.

"When you are training a dog, you may withhold a treat when it does something wrong. So no feedback means it did something wrong. On the other hand, when professors are grading tests, they may only mark wrong answers, so no feedback means you did something right," explained Matthew Taylor, Allred Distinguished Professor in the school of electrical engineering and computer science.

The researchers have begun working with physical robots as well as virtual ones.

They hope to eventually use the programme to help people learn to be more effective animal trainers.

The researchers recently presented their work at the international Autonomous Agents and Multiagent Systems conference in Singapore.​