Techniques are described to implement a probe sensor that improves data capture and data analysis. A probe sensor can be emulated in a virtual environment. A robot simulation session is initialized. The session includes a virtual environment with several objects and a set of robots. Each robot has a virtual sensor. A separate client controls each robot. Data perceived by the virtual sensor is provided to the client for controlling the robot. To capture the data the virtual sensor emits a plurality of rays, each ray transmitted in a stochastically selected direction, and performs raytracing to determine an object(s) in the virtual environment on which each ray is incident. The stochastic data capture can also be performed by a sensor in a real world scenario. Further, in some cases, the data captured by a sensor is stochastically sampled to improve the computing.

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for sharing learned information among robots. In some implementations, a robot obtains sensor data indicating characteristics of an object. The robot determines a classification for the object and generates an embedding for the object using a machine learning model stored by the robot. The robot stores the generated embedding and data indicating the classification for the object. The robot sends the generated embedding and the data indicating the classification to a server system. The robot receives, from the server system, an embedding generated by a second robot and a corresponding classification. The robot stores the received embedding and the corresponding classification in the local cache of the robot. The robot may then use the information in the cache to identify objects.

System includes: an operation terminal that (i) receives an operation instruction that is given by the operator to a movable section of a machine and (ii) senses an operation standby state that allows the operation instruction to be received; an image sensor estimating section configured to estimate a positional relationship between the operator and the machine; a model generating section configured to, in response to sensing of the operation standby state by the operation terminal, generate an operating direction indicating image; and in accordance with the positional relationship, in a direction that is in accordance with a direction in which the operator views the machine, the operating direction indicating image indicating an operating direction of the movable section; an combining section configured to generate a combined image obtained by combining the operating direction indicating image with a captured image of the movable section that has been photographed.

A method for simulating a braking operation of a robot wherein a dynamic model of the robot is used to determine, for a given initial state of the robot, a final state range with a plurality of possible final states of the robot as a result of the simulated braking operation.

A robotic system for arranging packages at a destination in a specified arrangement. The robotic system processes incoming packages, stores the packages in a temporary storage area, executes a simulate function to generate or update a simulated stacking plan, determines the occurrence of a palletizing trigger, and places the packages on the pallet according to the simulated stacking plan upon determining the occurrence of the palletizing trigger. The palletizing trigger can be one of a time limit trigger, a uniform layer trigger, a storage capacity trigger, or receiving a placement initiation command.

A system and method for machine understanding, using program induction, includes a visual cognitive computer including a set of components designed to execute predetermined primitive functions. The method includes determining programs using a program induction engine that interfaces with the visual cognitive computer to discover programs using the predetermined primitive functions and/or executes the discovered programs based on an input.

Methods, systems, and apparatus, including computer-readable storage devices, for robot navigation using 2D and 3D path planning. In the disclosed method, a robot accesses map data indicating two-dimensional layout of objects in a space and evaluates candidate paths for the robot to traverse. In response to determining that the candidate paths do not include a collision-free path across the space for a two-dimensional profile of the robot, the robot evaluates a three-dimensional shape of the robot with respect to a three-dimensional shape of an object in the space. Based on the evaluation of the three-dimensional shapes, the robot determines a collision-free path to traverse through the space.

A teaching device constructs, in a virtual space, a virtual robot system in which a virtual 3D model of a robot and a virtual 3D model of a peripheral structure of the robot are arranged, and teaches a moving path of the robot. The teaching device includes an acquisition unit configured to acquire information about a geometric error between the virtual 3D models, and a correction unit configured to correct the moving path of the robot in accordance with the information acquired by the acquisition unit.

A simulation device includes circuitry configured to: execute a first simulation of a first control of a first machine, wherein the first control is executed by a first controller; execute a second simulation of a second control of a second machine, wherein the second control is executed by a second controller so that the second machine operates in collaboration with the first machine; and control progress of the first simulation and progress of the second simulation to maintain a simulated ratio of a progress speed of the first simulation and a progress speed of the second simulation to be equal to a ratio of a progress speed of the first control and a progress speed of the second control.

A robotic grasp generation technique for machine tending applications. Part and gripper geometry are provided as inputs, typically from CAD files. Gripper kinematics are also defined as an input. Preferred and prohibited grasp locations on the part may also be defined as inputs, to ensure that the computed grasp candidates enable the robot to load the part into a machining station such that the machining station can grasp a particular location on the part. An optimization solver is used to compute a quality grasp with stable surface contact between the part and the gripper, with no interference between the gripper and the part, and allowing for the preferred and prohibited grasp locations which were defined as inputs. All surfaces of the gripper fingers are considered for grasping and collision avoidance. A loop with random initialization is used to automatically compute many hundreds of diverse grasps for the part.