Back to results
US 2024/0286272 A1Published application
Robot teaching with scans in and out of robot workspace
AssigneeScalable Robotics Inc.
InventorsThomas Andrew FUHLBRIGGE, Carlos MARTINEZ, Gregory ROSSANO
Life of this patent
Estimating the term…
Abstract
A robot teaching system that uses 3D scanning, movable devices, and pose selecting devices, either in or outside the robot workspace, in order to create a robot instruction. The system may include a 3D scanning device, a user input device, and a computation device. The computation device executes instructions to cause the 3D scanning device to capture a 3D scan of a workpiece, record a spatial point, and determine best fit geometries.
Claims3
A robot teaching system, comprising:
- a 3D scanning device;
- a user input device;
- a computation device;
- and
- a storage device in communication with the computation device, the storage device storing instructions that when executed by the computation device results in a method comprising:
- causing the 3D scanning device to capture a first 3D scan of a workpiece;
- recording a first spatial point in response to user input via the user input device;
- determining at least one best-fit geometry of at least a portion of the first 3D scan;
- projecting the first spatial point to a first geometric point having a predefined relationship to an individual one of the best-fit geometries;
- defining a robot instruction based on the first geometric point;
- and
- using a machine learning model to predict a quality resulting from the robot instruction.
9 more elements
A robot teaching system, comprising:
- a 3D scanning device;
- a user input device;
- a computation device;
- and
- a storage device in communication with the computation device, the storage device storing instructions that when executed by the computation device results in a method comprising:
- causing the 3D scanning device to capture a 3D scan of a workpiece;
- determining one or more best-fit geometries of at least a portion of the 3D scan;
- selecting one or more of the best-fit geometries in response to user input via the user input device;
- selecting one or more robot parameters in response to user input via the user input device;
- defining a robot instruction based on the one or more selected best-fit geometries and the one or more selected robot parameters;
- and
- recording the results of the robot instruction to train a machine learning model.
9 more elements
A robot teaching system, comprising:
- a first 3D scanning device and a second 3D scanning device;
- a user-input device;
- a computation device;
- and
- a storage device in communication with the computation device, the storage device storing instructions that when executed by the computation device results in a method comprising:
- causing the first 3D scanning device to capture a first 3D scan of a workpiece;
- recording a first spatial point in response to user input via the user input device;
- determining at least one best-fit geometry of at least a portion of the first 3D scan;
- projecting the first spatial point to a first geometric point having a predefined relationship to the at least one best-fit geometry;
- display to the user the first geometric point in real time with the user input;
- causing the second 3D scanning device to capture a second 3D scan of a region of the workpiece to which the first spatial point had been proximal;
- updating information from first 3D scan with information from the second 3D scan to create a merged 3D scan;
- projecting the first spatial point to a second geometric point having a predefined relationship to an individual one of the best-fit geometries of the merged 3D scan;
- and defining a robot instruction based on the second geometric point.
11 more elements
Dependencies are inferred from the claim text.