Data processing: generic control systems or specific application – Specific application – apparatus or process – Robot control
Reexamination Certificate
2000-08-03
2002-08-13
Cuchlinski, Jr., William A. (Department: 3661)
Data processing: generic control systems or specific application
Specific application, apparatus or process
Robot control
C700S247000, C700S251000, C700S252000, C700S259000, C700S262000, C700S264000, C318S568100, C318S568110, C318S568140, C901S009000, C901S042000, C901S047000, C414S730000
Reexamination Certificate
active
06434449
ABSTRACT:
BACKGROUND OF INVENTION
1. Field of the Invention
The present invention relates generally to calibration of components of a robot-cell including an industrial robot (robot), the tool center point of the end-effector (TCP) attached to the robot's flange, and additionally the fixture or positioner holding the production part upon which the robot performs some operation. Further, a preferred embodiment of the present invention relates to the automatic/“on-line” calibration of a robot and its TCP based on the measurements recorded by a sensor after contact or interception with random points along the three-dimensional contour of the end-effector close to the TCP.
2. Description of the Prior Art
Systems which employ the current state of technology for calibration of a robot and its TCP consist of two basic types: (1) “target-based” systems which can identify the robot, the fixture and TCP parameters but require the operator to attach one or more measurement target(s) at or near the physical location of the TCP; and (2) “search/feedback” systems which identify only TCP parameters but do not require operator intervention.
“Target-based” systems generally offer the benefit of allowing identification of robot parameters as well as fixture and TCP parameters. Traditionally these systems include an external measurement system including, but not limited to, a laser interferometer, a photogrammetry system, a theodolite system, or an optomechanical system with one or more measurement cables. In order to acquire measurement data used in the calibration process, each of these systems requires attachment of a target—or in the case of the optomechanical systems an adaptor—to one or more points at or near the physical TCP to be identified. Cost of the external measurement system and the requirement for operator intervention are the primary reasons that “target-based” systems are not considered suitable for use as an automated/“on-line” solution for identification of robot and TCP parameters.
In terms of suitability for use as an automated calibration solution, “search/feedback” systems possess a clear advantage over “target-based” systems—they do not require attachment of targets or other modification of the end-effector itself—thus eliminating the need for operator intervention in the calibration process. These systems traditionally include a “low resolution” sensor—as low as “1 bit” (i.e. “on/off” or “binary”) as is the case in either an optical beam, a proximity switch, or an electric contact, for example. As the TCP either breaks the optical beam or makes contact with the proximity switch, the robot position reported by the controller is recorded for use in the TCP identification process. As a consequence, the speed at which the robot moves toward either the optical beam or the proximity switch is inversely related to the accuracy of the identification process (i.e. a very slow robot speed is required to record highly accurate robot position information at the moment the TCP breaks the optical beam, for example). Furthermore, even though these “search/feedback” systems do not require specific measurement targets (as do the “target-based” systems), they still require the end-effector to intercept or contact the measurement sensor at specific and pre-determined locations with respect to the TCP. For that reason, the robot needs to approach the sensor in a methodical manner in order to “search”—through feedback—for the appropriate locations along the end-effector for the measurement sensor to intercept or contact (also referred to as “profiling” of the end-effector).
Furthermore, these “search/feedback” systems often only identify TCP parameters—they do not identify robot-related parameters. Moreover, these “search/feedback” systems possess several significant limitations as they typically require: (1) an initial approximation of the TCP values; (2) significant integration with the robot controller to establish a feedback loop which causes the TCP, for example, to break the optical beam several times; and (3) a significant amount of time for the robot to drive the TCP, for example, to break the optical beam several times.
Finally, the resulting accuracy of the TCP parameters identified with the “search/feedback” systems depends directly upon the following factors: (1) the extent to which the robot itself is already properly “calibrated” (i.e. such “low resolution” methods rely on the robot reporting its absolute position accurately in order to perform calibration of the TCP with accuracy); and (2) typically also the extent to which the TCP maintains a specific (and assumed upfront) orientation relative to the optical beam, proximity switch, or electrical contact.
SUMMARY OF THE INVENTION
In contrast to the “target-based” and “search/feedback” systems described above, the present invention involves a method for identification of both robot and TCP parameters and additionally the location of the measurement sensor, thus also either the location of the stationary fixture or other parameters of the multi-axis positioner. Further, in a preferred embodiment of the present invention, the robot and end-effector can be recalibrated quickly during operation, even between cycles of a production robot program—without operator intervention. Moreover, the present invention eliminates several obstacles presented by some “search/feedback” systems as the present invention: (1) can eliminate the need for an initial approximation of the TCP values; (2) can eliminate the need for a complex feedback loop (i.e. it can be controller independent, thereby, also reducing the amount of time required to perform the complete process); and (3) can reduce restrictions upon the motion of the TCP (i.e. orientation of contact with “on/off” sensor). Finally, the present invention eliminates the dependency upon proper “mastering” of the robot as the “true” joint offsets (or even more robot parameters if needed) are identified in the process of automatically identifying these and other robot parameters.
The present invention achieves these advantages by providing a method and apparatus for calibrating the robot and end-effector together with the location of the measurement sensor if needed (and therefore potentially the fixture on which the sensor(s) is or are located) based upon a mathematical description of the 3D contour of the tip of the end-effector. Further, in contrast to the “search/feedback” systems, the present invention proposes as one embodiment the use of a “high-resolution” displacement sensor rather than a “low resolution” sensor (even as low as “binary” or “on/off”) such as an optical beam or a proximity switch for example. A “high-resolution” sensor allows the robot to stop literally anywhere within the sensor's measurement range without the need for robot controller feedback, in turn allowing higher robot speeds and thus shorter cycle times.
Once the 3D contour of the tip of the end-effector has been described relative to the location of the TCP, the calibration system allows identification of both robot and TCP parameters (together with the location of the measurement sensor if needed, and therefore potentially the fixture and/or positioner) with only two additional pieces of information: (1) a set of measurements recorded by the measurement sensor following contact or interception with several points along the contour of the tip of the end-effector; and (2) the corresponding robot program which caused the tip of the end-effector to make contact with or intercept the measurement sensor.
REFERENCES:
patent: 4727471 (1988-02-01), Driels et al.
patent: 4753569 (1988-06-01), Pryor
patent: 4842475 (1989-06-01), Driels
patent: 5079043 (1992-01-01), Lambert
patent: 5148591 (1992-09-01), Pryor
patent: 5297238 (1994-03-01), Wang et al.
patent: 5400638 (1995-03-01), Kim
patent: 5506682 (1996-04-01), Pryor
patent: 5602967 (1997-02-01), Pryor
patent: 5608847 (1997-03-01), Pryor
patent: 5910719 (1999-06-01), Thorne
patent: 6295708 (2001-10-01), Spaid et al.
patent: 6324444 (2001-11-01), Wakaizumi et al.
patent:
Cuchlinski Jr. William A.
Marc McDieunel
LandOfFree
Method and device for automated robot-cell calibration does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and device for automated robot-cell calibration, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and device for automated robot-cell calibration will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2938648