Data processing: generic control systems or specific application – Specific application – apparatus or process – Article handling
Reexamination Certificate
2000-12-04
2003-07-08
Walsh, Donald P. (Department: 3653)
Data processing: generic control systems or specific application
Specific application, apparatus or process
Article handling
C700S229000, C700S190000, C700S217000, C700S189000
Reexamination Certificate
active
06591160
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to semiconductor wafer handling and processing equipment, and in particular, to a method and apparatus for calibration of a wafer handling robot relative to a station in a semiconductor workpiece tool.
2. Description of the Related Art
The introduction of workpiece handling robots into the semiconductor wafer fabrication process represented a significant advance in automation over manual and early transfer equipment for moving wafers between various stations at a workpiece tool, such as process tools and/or workpiece storage and handling locations. It is an important feature of conventional workpiece handling robots to be able to quickly and precisely access a workpiece from a first position, deliver it to a new location having different X, Y and Z coordinates in cartesian space, and set it down without risk of damage to the workpiece.
In order to accomplish this, a typical robot includes a shaft mounted in a base for translation along a vertical axis. A first arm, or link, is rotatably mounted to an upper end of the shaft, and a second arm, or link, is rotationally mounted to the opposite end of the first arm. The workpiece handling robot further includes an end effector pivotally attached to the second arm for supporting the workpiece. Various motors are further provided, conventionally mounted in the base, for translating the shaft, and for rotating the first and second arms such that the end effector may be controllably maneuvered in three-dimensional space.
To ensure that the end effector of the workpiece handling robot is precisely located during handling of workpieces, the robot must be calibrated relative to a workstation to and from which the robot transfers the workpieces. Each time the workstation is moved, as for example when it is replaced or repaired, the calibration process must be performed to ensure proper positioning of the robot to the workstation. Often a single robot operates with a plurality of work stations within a workpiece tool. In this instance, not only must the robot be calibrated to work with each of the stations, but any time one of the stations is moved, for reasons explained hereinafter, not only must the robot be re-calibrated to operate with the moved station, but it often must be re-calibrated to operate with each of the workstations serviced by that robot.
The calibration, or “teaching”, of a robot allows the robot to commit to stored memory the location of the station(s) relative to the robot. In order to calibrate (or re-calibrate) the robot to a particular station, a rough approximation of the correct positioning between the robot and station is established by the operator. Thereafter, the end effector is moved to the position on the workstation which the operator estimates is the proper position of the end effector to drop off and acquire workpieces to and from the station. Typically, the operator visually aligns an end effector with the center position of the workpiece supported in a cassette on the station. The operator may use a remote control “teach pendant” to position the end effector, or the operator may manually drag the end effector into what the operator estimates to be the proper position. This relative position between the robot and station is then stored in memory.
Conventional calibration procedures have many disadvantages. Visually aligning the end effector with the center of the workpiece has proven to be inaccurate and subjective because it is dependent on the accuracy of each individual operator. Furthermore, the structure of the cassette largely prevents visual inspection of a workpiece therein to determine its center. The operator in most cases only has visual access through a front of the cassette through which the workpieces are transferred. Moreover, in the case where the operator manually drags the arm of the robot, additional forces are placed on the robot which can lead to later difficulties with the functioning of the robot.
In addition, the calibration procedures have not been repeatable because there are no standard procedures for calibration of a workpiece handling robot to a station. One problem has been that the calibration procedure used by one operator to calibrate a workpiece handling robot to a station may be different than the calibration procedure used by a second operator calibrating the same robot to the same station. For example, a first operator may align a first point on the robot with the center of the workpiece, while a second operator uses a second, different point. This could result in faulty workpiece transfer to an existing station, even though the robot is properly calibrated to the new or modified station. Thus, as discussed above, when one station of a plurality of stations serviced by a single robot is added or otherwise moved, not only is it necessary to calibrate that particular station, but in many situations, each of the other stations must also be re-calibrated to ensure that each of the stations serviced by the robot are calibrated in the same way and off of the same reference points.
A still further problem has been that the calibration procedure used by the operator at one station may be different than the calibration procedure used by the same operator at a second station. As a result of these inconsistencies, one station may significantly out perform another station or cause damage to a workpiece. Since there is no repeatable and consistent calibration procedure, the reason for a particular performance at a station or for damage caused to a workpiece is not determinable.
Moreover, the calibration procedures have not been quantifiable because there is no way to compare the quality of the calibration procedure used by one operator to the quality of the calibration procedure used by a second operator. Similarly, there is no way to compare the quality of the calibration procedure used at one workstation to the quality of the calibration procedure used at a second workstation.
In an effort to deal with this problem, a known calibration method includes a sensor which is installed at every station in a workpiece tool. The sensor may be mounted at any position on the station which is reachable by the end effector so long as the location of the sensor does not interfere with the processing operations at the station. The calibration procedure is completed by having the robot seek the sensor in order to identify the location of the workpiece. This method has the advantage of eliminating errors due to operator inaccuracies. However, if there are a plurality of stations, each sensor must be mounted and calibrated with respect to the robot. For example, if there are twelve stations, then twelve sensors must be mounted and calibrated with respect to the wafer handling robot, or one sensor must be moved around manually to each of the twelve stations. Therefore, more sensors means the calibration system will be more expensive and time consuming.
In one known system, a support structure including a sensor is mounted directly to a work station in a know position relative to the work station. The robot is then coarsely aligned over the station, and is moved around so that the sensor can identify one or more outer edges of the robot. Once the outer edges of the robot have been identified, the position of the robot relative to the sensor and station can be determined and stored. A drawback to this type of system is that the sensor structure must either be mounted to each station, or a single sensor structure must be moved between each station, in order to align the robot to each station in the tool. Additionally, the sensor and control system are configured to identify the edges of an end effector of a single, known configuration. However, there is not a single uniform shape to end effectors, and thus the sensor structure has limited use.
SUMMARY OF THE INVENTION
It is therefore an advantage of the present invention to provide a robot calibration system which allows for determination of a position and orientation
Hine Graham L.
Hine Roger G.
Asyst Technologies, Inc.
Butler Michael E.
O'Melveny & Myers LLP
Walsh Donald P.
LandOfFree
Self teaching robot does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Self teaching robot, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Self teaching robot will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3010688