FANUC Open Platform
AI Robots Imitating Human Tasks
Two robot arms utilize imitation learning to acquire human skills and successfully fold a T-shirt—a highly flexible object.
Data Collection through Teleoperation
A human manually operates a leader robot to teleoperate a follower robot, performing the task of folding a T-shirt. Cameras installed in the workspace and on the robot grippers record paired datasets of robot motion and visual data.
Building an AI Model Through Imitation Learning
The collected dataset is annotated with a language instruction label, such as "fold the clothes." This data is then used to perform imitation learning on a Vision-Language-Action (VLA) foundation model—a highly advanced AI model—thereby constructing a specialized VLA model.
Autonomously Folding T-shirts
The VLA model takes visual data and language instructions as inputs, inferring optimal robot actions in real-time to generate precise robot motions. This enables the folding of flexible objects like T-shirts, whose shapes constantly change during the operation.
