Jenga robot: MIT unveils machine with new skills

The researchers at MIT are continually working on advances in robotics, and they are now showing off a new robot that combines vision and touch to teach itself to play the game of Jenga. The complex tasks that the robot has to perform in combination with the controlling AI provide a basic understanding of physics in the system that can also be used for other things.

The researchers concluded that compared to the other three state-of-the-art learning paradigms, including neural network and reinforcement learning, that the robot was fastest at reaching a certain number of successful block extractions (within 100 games) using this new approach. Players remove one block at a time from a tower comprised of 54 wooden blocks.

Alberto Rodriguez, the Walter Henry Gale Career Development Assistant Professor in the Department of Mechanical Engineering at MIT, pointed out that the robot stands out from previous milestones where robots could participate in cognitive games such as chess or Go.

Combining interactive perception and manipulation - whereby the robot would touch the tower to learn how and when to move blocks - is extremely hard to simulate and therefore the robot has to learn in the real world, he added.

Researchers at the Massachusetts Institute of Technology have created a program that can train a robotic arm to play Jenga, which requires physical interaction but also requires perception data from both touch and vision.

Helicopter carrying Nigeria's Vice President has crashed
In a series of tweets hours later, the APC thanked God the Vice President was safe and sound along with his aides. Concise News understands that the chopper was conveying Osinbajo to a campaign rally in Kogi State.

It could be used to sort recyclables from landfill trash, or assemble consumer products such as cell phones, Rodriguez said.

This research was supported, in part, by the National Science Foundation through the National Robotics Initiative. For each data cluster, the robot developed a simple model to predict a block's behavior given its current visual and tactile measurements. It then exerted a small amount of force in an attempt to push the block out of the tower.

In a video, the robot proves that it has actually already mastered the game quite well. "As a outcome, these systems require far more training data than humans do to learn new models or new tasks, and they generalize much less broadly and less robustly".

"There are many tasks that we do with our hands where the feeling of doing it "the right way" comes in the language of forces and tactile cues", Rodriguez says.

'For tasks like these, a similar approach to ours could figure it out'. The team found the success rate of the robot in keeping the tower upright while removing the wooden blocks was nearly on a par with that of human players.