Methods and apparatus to train interdependent autonomous machines are disclosed. An example method includes performing an action of a first sub-task of a collaborative task with a first collaborative robot in a robotic cell while a second collaborative robot operates in the robotic cell according to a first recorded action of the second collaborative robot, the first recorded action of the second collaborative robot recorded while a second robot controller associated with the second collaborative robot is trained to control the second collaborative robot to perform a second sub-task of the collaborative task, and training a first robot controller associated with the first collaborative robot based at least on a sensing of an interaction of the first collaborative robot with the second collaborative robot while the action of the first sub-task is performed by the first collaborative robot and the second collaborative robot operates according to the first recorded action.

System and method of learning and executing mutually coordinated paths of robot manipulators, including: manually guiding a first reference point of a first robot manipulator over a desired first path, acquiring the first path or acquiring a first set of poses for the first path and storing the first path or the first set of poses in a first data set, automatically traveling along the first path according to the first data set, while automatically traveling along the first path, manually guiding a second reference point of a second robot manipulator over a desired second path, acquiring the second path or acquiring a second set of poses for the second path and storing the second path or the second set of poses in a second data set, wherein the second data set is assigned to the first data set so that a location of the first path is at least approximately assigned to each location of the second path, and traveling along the first path by the first robot manipulator according to the first data set synchronized with traveling along the second path by the second robot manipulator according to the second data set.

Systems and a method are provided for programming a cobot for a plurality of cells of an industrial environment. A physical cobot is provided within a lab cell comprising physical lab objects. A virtual simulation system receives information inputs on a virtual cobot representing the physical cobot, regarding a virtual lab cell comprising virtual lab objects, and on a plurality of virtual industrial cells comprising virtual industrial objects. Inputs are received from the physical cobot's movement during teaching whereby the physical cobot is moved in the lab cell to the desired position(s) while providing, via a user interface, a visualization of the virtual cobot's movement within a meta cell generated by superimposing the plurality of virtual industrial cells with the virtual lab cell, so that collisions with any object are minimized. A robotic program is generated based on the received inputs of the physical cobot's movement.

Systems and a method for programming for a plurality of cells of an industrial environment. A physical cobot is provided within a lab cell comprising lab physical objects. A virtual simulation system with a user interface is provided. The virtual simulation system receives information inputs on the virtual cobot, on the virtual lab cell comprising lab virtual objects, and on a plurality of virtual industrial cells comprising virtual industrial objects. The virtual cobot and the physical cobot are connected together. A superimposed meta-cell is generated by superimposing the plurality of virtual cells and the virtual lab cell so as to obtain a single superimposed meta cell including a set of superimposed virtual objects. The virtual cobot is positioned in the superimposed meta cell. Inputs are received from the physical cobot's movement during teaching whereby the physical cobot is moved in the lab cell to the desired position(s) while providing, via the user interface, a visualization of the virtual cobot's movement within the superimposed meta cell so that collisions with any object are minimized. A robotic program is generated based on the received inputs of the physical cobot's movement.

Methods and apparatus to train interdependent autonomous machines are disclosed. An example method includes performing an action of a first sub-task of a collaborative task with a first collaborative robot in a robotic cell while a second collaborative robot operates in the robotic cell according to a first recorded action of the second collaborative robot, the first recorded action of the second collaborative robot recorded while a second robot controller associated with the second collaborative robot is trained to control the second collaborative robot to perform a second sub-task of the collaborative task, and training a first robot controller associated with the first collaborative robot based at least on a sensing of an interaction of the first collaborative robot with the second collaborative robot while the action of the first sub-task is performed by the first collaborative robot and the second collaborative robot operates according to the first recorded action.

A robot system includes a robot including a robot arm, and a first hand and a second hand which are connected to the robot arm and which are provided to independently rotate about an axis on the robot arm; and a controller configured to control an operation of the robot. When the robot arm and the first hand are operated so that the first hand reaches a predetermined target position, teaching values for the first hand in the target position is generated. When the first hand and the second hand are rotated based on the teaching values for the first hand, a relative error in rotation amount around the axis between the first hand and the second hand is acquired and stored in a memory. Teaching values for the second hand is generated from the teaching values for the first hand based on the acquired relative error.