The invention relates to a process for painting a workpiece using a painting robot including a robot arm equipped with a paint spraying device, the process including, an operation S1 of modeling a realistic 3D model corresponding to the workpiece as deformed and positioned in a paint cell, the realistic 3D model including paint trajectory information suitable for the workpiece as deformed and positioned in the paint cell, and a paint spraying operation S2 during which the paint spraying device is moved along the paint trajectory opposite the workpiece.

The present disclosure provides is a method and device for accurately estimating a pose of a charging socket of an electric vehicle regardless of a shape of the charging socket, so that an electric vehicle charging robot may precisely move a charging connector toward the charging socket of the electric vehicle and couple the charging connector to the charging socket. According to an aspect of an exemplary embodiment, a method of estimating the pose of the charging socket of an electric vehicle includes: acquiring an RGB image and a depth map of the charging socket; detecting a keypoint of the charging socket based on the RGB image; deriving a first estimated pose of the charging socket based on the depth map; and deriving a second estimated pose of the charging socket based on the keypoint of the charging socket and the first estimated pose.

Techniques are described herein for robotic control using value distributions. In various implementations, as part of performing a robotic task, state data associated with the robot in an environment may be generated based at least in part on vision data captured by a vision component of the robot. A plurality of candidate actions may be sampled, e.g., from continuous action space. A trained critic neural network model that represents a learned value function may be used to process a plurality of state-action pairs to generate a corresponding plurality of value distributions. Each state-action pair may include the state data and one of the plurality of sampled candidate actions. The state-action pair corresponding to the value distribution that satisfies one or more criteria may be selected from the plurality of state-action pairs. The robot may then be controlled to implement the sampled candidate action of the selected state-action pair.

A system, a device and a method for controlling a physical or chemical process. The method includes: determining a second a posteriori model based on a first a posteriori model that describes the relationship between an input variable and an output variable of a process related to the physical/chemical process, the second a posteriori model describing the relationship between an input variable and an output variable of the physical or chemical process; and controlling the physical or chemical process using the second a posteriori model.

Example implementations described herein can involve a plurality of data repositories involving a data repository configured to manage data versions of data sets corresponding to robot simulation versions; a code repository configured to manage code versions of executable code corresponding to the robot simulation versions; and a robot model repository configured to manage model versions of robot models corresponding to the robot simulation version. Responsive to a request of execution of a robot simulation, fetch, from the plurality of data repositories, corresponding one or more of the data sets having a data version from the data versions that corresponds to a robot simulation version of the robot simulation from the robot simulation versions, corresponding executable code, and a corresponding robot model.

A method of configuring robot fleets with additive manufacturing capabilities includes receiving a request for a robotic fleet to perform a job and determining a job definition data structure based on the request. The job definition data structure defines a set of tasks to be performed in furtherance of the job. The method includes determining a provisioning configuration for each additive manufacturing system based on the task to which the additive manufacturing system is assigned, the set of 3D printing requirements, the printing instructions, and the status of the additive manufacturing system. The method includes provisioning the additive manufacturing system based on the provisioning configuration and a set of additive manufacturing system provisioning rules that are accessible to an intelligence layer to ensure that provisioned systems comply with the provisioning rules. The method includes deploying the robotic fleet based on the robotic fleet configuration data structure to perform the job.

Embodiments of the disclosure provide a robot control method, apparatus and device, a computer storage medium and a computer program product and relate to the technical field of artificial intelligence. The method includes: acquiring environment interaction data and an actual target value, indicating a target that is actually reached by executing an action corresponding to action data in the environment interaction data; determining a return value after executing the action according to state data, action data and the actual target value at the first time of two adjacent times; updating a return value in the environment interaction data by using the return value after executing the action; training an agent corresponding to a robot control network by using the updated environment interaction data, and controlling the action of a target robot by using the trained agent.

A vacuum tube assembly for removing material is disclosed, including: a vacuum generator configured to generate a vacuum airflow; one or more tubes coupled to the vacuum generator and configured to channel the vacuum airflow; and an actuation mechanism coupled to the one or more tubes, wherein the actuation mechanism is configured to actuate at least one tube from a first position relative to a material stream to a second position relative to the material stream, wherein the first position is farther from the material stream than the second position.

One variation of a method for autonomously scanning and processing a part includes: collecting a set of images depicting a part positioned within a work zone adjacent a robotic system; assembling the set of images into a part model representing the part. The method includes segmenting areas of the part model—delineated by local radii of curvature, edges, or color boundaries—into target zones for processing by the robotic system and exclusion zones avoided by the robotic system. The method includes: projecting a set of keypoints onto the target zone of part model defining positions, orientations, and target forces of a sanding head applied at locations on the part model; assembling the set of keypoints into a toolpath and projecting the toolpath onto the target zone of the part model; and transmitting the toolpath to a robotic system to execute the toolpath on the part within the work zone.

A method including acquiring measurements representing torque applied to a connection between a first tubular and a second tubular, a rotational position of the first tubular relative to the second tubular, or both, obtaining a plurality of make-up parameters representing conditions under which the connection is fully made, generating a plurality of rules for connection evaluation based on the make-up parameters, and automatically evaluating the connection using a computer. Automatically evaluating includes applying machine learning, the plurality of rules, or a combination thereof to a dataset of the measurements. The method also includes recommending accepting or rejecting the connection based on the automatic evaluation.