Robot Learns to Play a Mean Game of Jenga And it is Incredible At It


Alberto Rodriguez, assistant professor of mechanical engineering at MIT joined with a team from the MIT MCube Lab produced a very accurately moving robot that uses machine learning hardware to play a mean game of Jenga. The robot smoothly removes a wooden piece in two movements and the place it at the top of the pile and is able to replicate this movement many times while learning and retaining information about the conditions in real time.

The robot efficiently and clearly identifies when a piece feels stuck or free and decides how to extract it using far less data. This approach as a successful example of AI moving into the physical world. The robot learns as it interacts with its environment and captures some of the essential skills that define human manipulation.

via MIT News


Leave a Reply