Robot calibration for AR and digital twin
A method and system for calibration of an augmented reality (AR) device's position and orientation based on a robot's positional configuration. A conventional visual calibration target is not required for AR device calibration. Instead, the robot itself, in any pose, is used as a three dimensional (3D) calibration target. The AR system is provided with a CAD model of the entire robot to use as a reference frame, and 3D models of the individual robot arms are combined into a single object model based on joint positions known from the robot controller. The 3D surface model of the entire robot in the current pose is then used for visual calibration of the AR system by analyzing images from the AR device camera in comparison to the surface model of the robot in the current pose. The technique is applicable to initial AR device calibration and to ongoing device tracking.
1. A method for calibrating an augmented reality (AR) device to a robot base reference frame, said method comprising:
providing a three-dimensional (3D) model of each structural component of a robot to the AR device;
providing position data for all joints of the robot in a current pose from a robot controller to the AR device;
computing a 3D model of the robot in the current pose, by the AR device, using the 3D models of the components and the position data;
capturing image or sensor data of the robot in the current pose by one or more camera or sensor in the AR device; and
comparing the image or sensor data of the robot to the 3D model of the robot to determine a position and orientation of the AR device with respect to the robot base reference frame.
2. The method according to claim 1 wherein the 3D models of each of the structural components of the robot are computer aided design (CAD) solid or surface models.
3. The method according to claim 1 wherein the AR device is a headset apparatus worn by an operator, or the AR device is a smart phone, a tablet computing device or a robot teach pendant held by the operator.
4. The method according to claim 1 wherein the robot is an articulated robot having multiple arms connected in series, and the position data is joint angular position data measured by joint encoders in the robot.
5. The method according to claim 1 wherein the robot is an articulated robot having multiple arms connected in series, and the position data is joint Cartesian coordinates measured by sensors on the robot.
6. The method according to claim 1 wherein the robot is a delta-type robot having multiple parallel links connecting a moving pick-and-place head to a fixed hub.
7. The method according to claim 1 wherein the image or sensor data is digital images of the robot taken by one or more cameras in the AR device.
8. The method according to claim 1 wherein the image or sensor data is data from a laser sensor, an audio spectrum sensor, a time-of-flight based sensor, a radar sensor or a LiDar sensor.
9. The method according to claim 1 further comprising continuously recalculating the position and orientation of the AR device with respect to the robot base reference frame, including making inertial odometry calculations based on signals from accelerometers and yaw rate sensors in the AR device.
10. The method according to claim 9 further comprising error-correcting the inertial odometry calculations by comparing image or sensor data of the robot in a new pose to a 3D model of the robot in the new pose to determine a position and orientation of the AR device with respect to the robot base reference frame.
11. The method according to claim 1 further comprising using the position and orientation of the AR device with respect to the robot base reference frame, by an AR application program running on the AR device, in order to properly depict virtual elements on a display of the AR device.
12. The method according to claim 1 wherein providing the position data, computing the 3D model of the robot, capturing the image or sensor data and comparing the image or sensor data to the 3D model occur while the robot is parked in a stationary position.
13. The method according to claim 1 wherein providing the position data, computing the 3D model of the robot, capturing the image or sensor data and comparing the image or sensor data to the 3D model occur while the robot is running a motion program.
14. A system for calibrating an augmented reality (AR) device to a robot base reference frame, said system comprising:
an industrial robot having a plurality of structural components;
a robot controller in communication with the robot, said controller including a processor and memory and being configured with robot operational control software; and
an augmented reality (AR) device in communication with the robot controller, said AR device having one or more camera or object sensor, position tracking sensors, a display, and a processor and memory configured to run an AR application,
where the robot controller provides position data for all joints of the robot in a current pose to the AR device, and
the AR device computes a 3D model of the robot in the current pose using 3D models of the structural components and the position data, captures image or sensor data of the robot in the current pose using the one or more camera or sensor, and compares the image or sensor data of the robot to the 3D model of the robot to determine a position and orientation of the AR device with respect to the robot base reference frame.
15. The system according to claim 14 wherein the AR device is a headset apparatus worn by an operator, or the AR device is a smart phone, a tablet computing device or a robot teach pendant held by the operator.
16. The system according to claim 14 wherein the robot is an articulated robot having multiple arms connected in series, or the robot is a delta-type robot having multiple parallel links connecting a moving pick-and-place head to a fixed hub.
17. The system according to claim 14 wherein the image or sensor data is digital images of the robot taken by one or more cameras in the AR device, or the image or sensor data is object data from a laser sensor, an audio spectrum sensor, a time-of-flight based sensor, a radar sensor or a LiDar sensor.
18. The system according to claim 14 wherein the 3D models of the structural components of the robot are computer aided design (CAD) solid or surface models.
19. The system according to claim 14 wherein the AR device continuously recalculates the position and orientation of the AR device with respect to the robot base reference frame, including making inertial odometry calculations based on signals from the position tracking sensors, and error-correcting the inertial odometry calculations by comparing image or sensor data of the robot in a new pose to a 3D model of the robot in the new pose.
20. The system according to claim 14 wherein the position and orientation of the AR device with respect to the robot base reference frame are used in an AR application program running on the AR device in order to properly depict virtual elements on a display of the AR device.
21. The system according to claim 14 wherein providing the position data, computing the 3D model of the robot, capturing the image or sensor data and comparing the image or sensor data to the 3D model occur while the robot is parked in a stationary position or occur while the robot is running a motion program.
22. A method for calibrating an augmented reality (AR) device to a machine base reference frame, said method comprising:
providing a three-dimensional (3D) model of each structural component of an articulated machine to the AR device;
providing position data for all parts of the machine in a current pose from a machine controller to the AR device;
computing a 3D model of the machine in the current pose, by the AR device, using the 3D models of the components and the position data;
capturing image or sensor data of the machine in the current pose by one or more camera or sensor in the AR device; and
comparing the image or sensor data of the machine to the 3D model of the machine to determine a position and orientation of the AR device with respect to the machine base reference frame.
23. The method according to claim 22 wherein the position and orientation of the AR device with respect to the machine base reference frame are used in a program running on a simulation system in order to properly display movements of a digital twin of the machine, where the digital twin is a virtual replica of the machine which is calibrated by the AR device.
24. The method according to claim 22 further comprising mapping locations and orientations of other machines and other objects in the machine base reference frame, by the AR device, including comparing images of the other machines and other objects to known solid/surface models of the other machines and other objects.