Abstract: An exemplary method and system are disclosed to flexibly and adaptably manufacture and assemble a workpiece by using recordings of a user in machine learning/artificial intelligence algorithms to train a robot for subsequent automated manufacture. Machine learning and artificial intelligence learning can generate libraries of generalized dynamic motion primitives that can be subsequently combined for any type of manufacturing or assembling activity. The exemplary method and system can flexibly generate a model of an existing workpiece as a template or primer workpiece that can then be used in conjunction with the DMP operations to fabricate subsequent workpieces.

Abstract: A robot gripper includes two fingers of a grasper assembly configured to perform grasping motions via actuation of independent cable ends of a plurality of cables, and configured to move toward or away from each other to perform the grasping motion. Wherein each gripper finger is actuated by a pair of cables, a cable of the pair slides in a flexible sheath when actuated by a motor, moving the gripper finger in an opposite direction of another cable of the pair also in a flexible sheath, providing equal motions of each cable in the pair in opposite directions. A motor assembly including the motors is mounted at a location separate from the grasper assembly with the flexible sheathing extending between the assemblies. Such that the separate assembly mounting arrangement provides an improved ratio between a gripping force of the grippers versus the robot-lifted mass of the grasper assembly.

Abstract: A system includes a robot having a module that includes a function for mapping natural language commands of varying complexities to reward functions at different levels of abstraction within a hierarchical planning framework, the function including using a deep neural network language model that learns how to map the natural language commands to reward functions at an appropriate level of the hierarchical planning framework.

Abstract: A method includes enabling a robot to learn a mapping between English language commands and Linear Temporal Logic (LTL) expressions, wherein neural sequence-to-sequence learning models are employed to infer a LTL sequence corresponding to a given natural language command.

Abstract: A robot gripper, including at least two gripper fingers of a grasper assembly configured to perform grasping motions via actuation of independent cable ends of a plurality of cables, and configured to move toward or away from each other to perform the grasping motion. Wherein each gripper finger is actuated by a pair of cables, a cable of the pair slides in a flexible sheath when actuated by a motor, moving the gripper finger in an opposite direction of an other cable of the pair also in a flexible sheath, providing equal motions of each cable in the pair in opposite directions. A motor assembly including the motors is mounted at a location separate from the grasper assembly with the flexible sheathing extending between the assemblies. Such that the separate assembly mounting arrangement provides an improved ratio between a gripping force of the grippers versus the robot-lifted mass of the grasper assembly.

Abstract: A system includes a robot having a module that includes a function for mapping natural language commands of varying complexities to reward functions at different levels of abstraction within a hierarchical planning framework, the function including using a deep neural network language model that learns how to map the natural language commands to reward functions at an appropriate level of the hierarchical planning framework.

Abstract: A system includes a robot having a module that includes a function for mapping natural language commands of varying complexities to reward functions at different levels of abstraction within a hierarchical planning framework, the function including using a deep neural network language model that learns how to map the natural language commands to reward functions at an appropriate level of the hierarchical planning framework.

Abstract: A method includes enabling a robot to learn a mapping between English language commands and Linear Temporal Logic (LTL) expressions, wherein neural sequence-to-sequence learning models are employed to infer a LTL sequence corresponding to a given natural language command.

Abstract: A system includes a robot having a module that includes a function for mapping natural language commands of varying complexities to reward functions at different levels of abstraction within a hierarchical planning framework, the function including using a deep neural network language model that learns how to map the natural language commands to reward functions at an appropriate level of the hierarchical planning framework.