Jenga is one of the most popular indoor games. A slow-moving game really tests your patience and focus. Soon, humans will have to compete with robots in the Jenga.
Rather than relying solely on visual information, players have to poke, tap, and experience on individual wooden blocks to choose which to remove from the tower. But thanks to the machine learning algorithm, MIT researchers were able to teach robots how to successfully play Jenga – just the basic instructions – effective conquest for tactile robotics.
In the research paper published today by the journal Science Robotics, the robot’s approach to complete the status of each block is entirely outlined. Next, following the block condition, the next action is described for “successful extraction” of pieces by predicting a block’s future state.
It can be either a push or pull a piece, a single millimeter at a time. If the Force sensors help to analyze the situation continues to figure out if something is wrong or if the tower is likely to collapse.
Senior author Alberto Rodriguez said a popular science that the robot can learn from past mistakes and adjusting its behavior after the tower collapses by “building nuggets of experience.” In other words, he knows what a successful footprint is “feels” like.
The robot plays Jenga in a similar way that how humans play: We have brought strategies – refrain from a collapse of the tower in the process – Imagine the result – and think which piece to remove through feels.
This machine is an excellent and fun example of learning power. But the researchers’ robot will not win any major Jenga competition.
Rodriguez said in popular science, It’s “good enough so that it could play against a human,” but won’t “achieve superhuman performance.”