Apparatus, systems, methods, and articles of manufacture for automatic robot perception programming by imitation learning are disclosed. An example apparatus includes a percept mapper to identify a first percept and a second percept from data gathered from a demonstration of a task and an entropy encoder to calculate a first saliency of the first percept and a second saliency of the second percept. The example apparatus also includes a trajectory mapper to map a trajectory based on the first percept and the second percept, the first percept skewed based on the first saliency, the second percept skewed based on the second saliency. In addition, the example apparatus includes a probabilistic encoder to determine a plurality of variations of the trajectory and create a collection of trajectories including the trajectory and the variations of the trajectory. The example apparatus also includes an assemble network to imitate an action based on a first simulated signal from a first neural network of a first modality and a second simulated signal from a second neural network of a second modality, the action representative of a perceptual skill.

In the present disclosure, methods, systems, and apparatuses for in-process assembly error correction are described. In various embodiments, a target arrangement of parts of an assembly may be obtained, with the target arrangement including a first target position of a first part, a second target position of a second part, and a third target position of a third part. The first part and the second part may be robotically joined based on the first target position and the second target position to obtain a first subassembly of the assembly, with the first subassembly having a first physical arrangement that includes the physical arrangement of the first and second parts after joining. The first physical arrangement may be fitted to the target arrangement to obtain a fitted first physical arrangement. The first subassembly and the third part may be robotically joined based on the fitted first physical arrangement.

A system and method for programming a robot includes providing a 3D representation of workpieces to be handled by the robot, and of a working environment; synthesizing and displaying a view of the working environment comprising an image of the workpieces at respective initial positions; identifying matching features of the selected workpiece and of the working environment which are able to cooperate to hold the workpiece in a final position in the working environment, and a skill by which the matching features can be brought to cooperate; identifying an intermediate position from where applying the skill to the workpiece moves the workpiece to the final position; and adding to a motion program for the robot a routine for moving the workpiece from its initial position to the intermediate position and for applying the skill to the workpiece at the intermediate position.

A system for assembling a vehicle platform includes a robotic assembly system having at least two robotic arms, a vision system capturing images of an assembly frame, and a control system configured to control the robotic assembly system to assemble the vehicle platform based on images from the vision system, force feedback from the at least two robotic arms, and a component location model. The control system is further configured to identify assembly features of a first component and a second component of the vehicle platform from the images, operate the robotic arms to orient the first component and the second component to respective nominal positions based on the images and the component location model, and operate the robotic arms to assemble the first component to the second component based on the force feedback.

Provided is an automotive glass setting apparatus, including a setting base supporting a glass having a plurality of edges, a plurality of alignment units allowing the glass to be aligned with the setting base; a scanning unit scanning the edges of the glass, a plurality of moving mechanisms moving the plurality of alignment units, and a controller controlling the moving mechanisms based on data scanned by the scanning unit to align a center of the glass with a center of the setting base.

An approach to positioning one or more robotic arms in an assembly system may be described herein. For example, an apparatus may include a first robotic arm having a distal end and a proximal end. The distal end may be configured for movement and the proximal end may secure the first robotic arm. The apparatus may further include a camera connected with the distal end of the first robotic arm. The camera may be configured to capture image data of a marker connected with a second robotic arm and provide the image data to a computer. The computer may generate a set of instructions for the first robotic arm based on the image data of the marker. The movement of the first robotic arm may be caused by the computer according to the generated set of instructions.

A robotic system is provided for assembling parts together. In the assembly process, both parts are moving separately with one part moving on an assembly base and another part moving on a moveable arm of a robot base. Motion data is measured by an inertial measurement unit (IMU) sensor. Movement of the robot base or moveable arm is then compensated based on the measured motion to align the first and second parts with each other and assemble the parts together.

Methods and apparatuses for performing automated operations using a high-density robotic cell. An apparatus comprises a first plurality of robotic devices; a second plurality of robotic devices; and a control system. Each of the second plurality of robotic devices is coupled to a single function end effector. The control system controls the second plurality of robotic devices to concurrently perform tasks at a plurality of locations on an assembly, while the first plurality of robotic devices independently maintain a clamp-up at each of the plurality of locations.

A robot control system includes a relative relationship calculating section configured to calculate a relative relationship between a first member and a second member at least at one of a plurality of points based on data acquired by a vision sensor, a contact point determination section configured to determine a contact point between the first member and the second member based on the calculated relative relationship, a control point setting section configured to set a control point based on the determined contact point, and a fitting control section configured to control fitting of the plurality of points based on the set control point.

A program generation apparatus according to one or more embodiments may extract, from a series of motions defined in a motion program, a motion to be corrected based on a difference in attribute between a first component indicated as a target of a component change and a second component to replace the first component, and generate a new motion program by correcting a command value of the extracted motion to compensate for the difference in the attribute.