[ad_1]
A group of MIT researchers has developed a robot that plays Jenga and promises to be unbeatable. It has a force sensor, a clamp and a camera. In this way, he sees and perceives the tower. Then he deals with these visual and sensory stimuli and decides which room to move.
Unlike other artificial intelligence systems, based on a large database, this robot uses a hierarchical model to extract parts accurately and carefully.
Divide the possible interactions between the robot and the Jenga Tower into groups, each governed by its own laws of physics. Thus, the robot identifies, clearly, when a part is locked or not and according to who makes a decision.
In turn, before removing a piece, compare the information that he perceives on his environment with the movements made previously. Learn in real time and improve performance minute by minute.
This robot is a successful example of the use of artificial intelligence on the physical plane. The system not only sees, but also perceives by the "touch" and learns in real time.
This gives the robot precision and traction, which could be transferred to other industries to allow, for example, the badembly of small parts.
Source link