DE19843602A1 - Processing unit movement detection for manufacturing process - Google Patents

Abstract

Spatial coordinates of each processing unit are continuously recorded by means of stationary cameras forming a camera array, and an evaluation unit. A translation vector describing the movement of the respective processing unit is determined for each spatial coordinate, and data corresponding to a spatial path are determined for each processing unit through stringing together of the determined translation vectors. The method involves detecting movements of processing units in a predetermined range during a manufacturing process. Spatial coordinates of each processing unit are continuously recorded by means of at least two stationary cameras arranged in the range, which form a camera array, and an evaluation unit. A translation vector describing the movement of the respective processing unit is determined for each spatial coordinate, and data corresponding to a spatial, preferably two-dimensional path are determined for each processing unit through stringing together of the determined translation vectors. An Independent claim is provided for a corresponding movement detection arrangement.

Description

The invention relates to a method and a device for detecting the movements of process units during of a production process in a given evaluation area.

For the control of material flows and production pro in a production facility and for an automatic er Quality and product data will always be recorded more important that individually transported units (batches, Hordes, etc.) have an identifier that they use in the pro production process can be clearly identified. This is all the more important when the to be finished Products cannot be distinguished from the outside. A Examples of this are wafers with different processes histories in wafer or chip production.

First, the wording is standardized the following definition of terms agreed:

All kinds of product, material, transportation and / or Equipment items, combinations or summaries solutions are called process units. Farther are all types of product, material and / or Equipment movements and / or flows as a pro designated unit movement.  

The description of the invention is clear only for the sake of chip production. This represents all types of products on processes in wafer, semiconductor, electronics, Micromechanics and / or precision engineering production.

Precisely because it is so important for chip production clear tracking and logging of mate Rial rivers have become more current State of the art in the semiconductor industry various Automatically or automatically evaluable or readable Labeling systems for indoor and outdoor use process units.

The simplest known technique is to use one Barcodes with the help of laser scanners or barcodes can be read.

A far more advanced labeling technology represent the so-called infrared tags (IR-ID) are label elements already in Form of an integrated chip. Is activated the chip through light exposure through a photo voltaic process. Data exchange takes place through an infra red transmission.

Also in the direction of an intelligent (chip) Marking that has become known recently the so-called transponders. These contain similar like the IR IDs a chip, but with the difference that using power and data transmission of radio waves happens. Your own energy source is also in these RF-IDs (RF = Radio Frequency) not mandatory.

A purely logical material is particularly prone to errors tracking.  

The main drawbacks of the prior art are in that the transponders (RF IDs), the infrared identifier drawings (IR IDs) and all other known ver systems due to the relatively short ranges and the limited selectivity (e.g. multiple processes units with transponders on a table) are the complete product, material or equipment movement without gaps, because all processes between the reading or identification points are fundamental invisible. It is precisely these gaps in coverage, which can lead to problems with the known systems nen.

The object of the invention is therefore to provide a method for Detect the movements of process units during a Production process in a given evaluation area rich to indicate at which the above Disadvantages are eliminated.

This task is accomplished by a procedure with the in claim 1 specified features and by a device solved with the features specified in claim 15. Before partial further training results from the dependent Claims.

The advantages of the invention are in particular that even the so-called blind spots or chaotic ones Intermediate steps or buffer storage between the individual Process steps or process sections can be detected nen. As blind spots or chaotic intermediate steps or buffer stores are especially those process steps or process unit movements meant the through the already known detection options (Transponder, barcode, infrared identification) not or can only be insufficiently recorded or tracked.  

The invention can be used both when the Pro Unit movement using a technical Device such as conveyor belt, conveyor, robot, etc., he follows and even when this process unit movement is carried out manually by the operator staff.

An interaction with the already known systems increases the overall safety of the movement performed version.

Advantageously, according to the invention, the Pro Unit movement through an optical detection system detected as a two- or three-dimensional space trace and abge saves or logs. This corresponds to an opti acquisition by a two- or three-dimensional Motion analysis. This makes it possible at any time not only the type in the specified evaluation range number of process units, but also their exact order to determine location.

Further advantageous properties of the invention result itself based on the explanation of exemplary embodiments of the figures.

It shows:

Fig. 1 is a schematic representation of a first example of the invention From guide,

Fig. 2 is a schematic representation of a second embodiment of the invention,

Fig. 3 is a block diagram of the evaluation unit 3 of Fig. 1,

Fig. 4 is a diagram illustrating the arrangement of the cameras in a clean room,

Fig. 5 is a sketch of the clean room having therein befind current process units,

Figure 6 units. A second drawing that illustrates a clean room with Prozeßein therein,

Fig. 7 is a sketch illustrating the determination of a three-dimensional space, and

Fig. 8 is a sketch illustrating a wide ren embodiment for the invention.

As is apparent from FIGS . 1 and 2, the claimed device has image recording units 1 and 5 , respectively, which are arranged in a suitable manner and are connected to an evaluation unit 3 and 7 , respectively. In the following, a composite of one or more image recording units is referred to as a camera array.

The cameras used can both image data in analog form (BAS signal, standard video signal) as even in already digitized, d. H. binary form lie remote.

Analog black and white video cameras with standard video signal (television technology) are now available on the market at very low cost and with an extremely small size (32 mm × 32 mm × 14 mm), which is very beneficial for the construction of a large camera array. The individual video signals 2 , as shown in FIG. 1, are individually guided to the evaluation unit 3 and brought into a digital form therein.

A bus variant, as shown in Fig. 2, is possible in principle with analog cameras, but requires that the cameras have a switch for the video signal integrated and adhere to a bus protocol (access control).

The digital camera technology, on the other hand, is currently a very undergoes rapid further development, opens the Mög the image data without an additional signal wall effort with a computer or other digi tal image processing unit to process.

Working with a bus or networking technology 6 ( FIG. 2) between the cameras 5 and the evaluation unit 7 is simple because the image data are already in digital form and can therefore be easily transmitted digitally. The bus or networking technology as well as the different protocol variants are sufficiently known in the field of automation and computer networks, so that an exhaustive presentation can be dispensed with at this point.

Representing all other possible solutions sets, such as the image data from one or more cameras get into an image evaluation unit, the structure is on hand explained in more detail an analog solution.

Fig. 3 shows this continuously enter the captured by the camera array sub-images to be monitored evaluation range (see also Fig. 4) the schematic structure for egg NEN evaluation unit 3 with eight analog inputs 2, on the. In this example, only one camera signal is selected by a multiplexer 8 and passed to the digitizing unit, the so-called frame grabber 9 . The image data are then available in binary form and can be processed digitally in the image evaluation unit 10 and in particular evaluated.

At 50 fields per second, the multiplexing of eight cameras results in approximately five fields per second. With an average walking speed of 1.6 m / s, there is a path difference of approx. 30 cm between two fields of a camera. For one transport operation each. B. by only one person (operator 18 ), this is sufficient for an evaluation.

Should several people work (high transport volume men) or simultaneously more complex process unit movements are analyzed and recorded, so is a higher frame rate required. You can do this simply by using less Ka meras per evaluation unit or by using di gital cameras.

A translational movement of a process unit, for example by being carried on by the operator, can be most easily recognized by comparing a pattern. This is particularly simple if the process unit is always viewed from the same perspective (e.g. upright), as seen in FIGS. 5, 6 and 7, element 27 .

However, it is particularly common in semiconductor production before that the process units in the process equipment on the Inserted side tilted so that, for example, the Wafers can be easily removed by a robotic arm can. By tilting the effort for Identification of the process unit in the picture but still remains easily solvable. You put in additional physical knowledge as with for example the knowledge of inverse kinematics, plausi balance considerations or anatomical specifications such as In Formations about the dimensions and restrictions of movement of the human body, can also be difficult picture situations or unclear picture data, such as B. a cover of the process unit through the body of the Operators, master.  

Also, individual process units that are still in the system are not known, d. H. their initial space trace or fro was not detected by the system, found and through appropriate measures such. B. a message the operator and / or identification by means of a conventional reading point.

The simplest approach of an image evaluation unit 10 with high computing power is to use a digital signal processor DSP. Such DSPs only make it possible to master the enormous computing effort for the image evaluations due to the constantly growing performance. However, such DSPs also reach their limits when it comes to the additional considerations mentioned above (physical knowledge and plausibility checks).

A new and even more powerful approach is to use an FPGA (Field Programmable Field Array) as an image evaluation unit 10. With suitable programming, you are not only able to process complex evaluation algorithms in parallel to a high degree, but also the entire control logic for to integrate the multiplexer, the frame grabber or for the control of digital cameras. A combination of FPGA and DSP is even more powerful, especially for 3D evaluations. The highest performance is of course achieved if, in extreme cases, each camera has its own evaluation unit, these evaluation units being logically connected to one another via a network 4 .

The position of a process unit, ie the coordinates in the two-dimensional image matrix, is first extracted by the image evaluation unit 10 . Rotations of the process unit around the vertical axis can also be obtained here by the position of the pattern comparison and used for further evaluations.

By comparing with the previous image, one obtains the shift caused by movement, ie the translation vectors. If one restricts himself to a two-dimensional evaluation, the space track, as illustrated in FIGS . 5, elements 19 and 6 , element 21 , is composed simply by stringing the translation vectors together.

The image evaluation unit 10 must of course take into account the perspectives of the individual cameras in the camera array and put the individual images together accordingly. Furthermore, the crossing of an image boundary must be taken into account arithmetically when entering the adjacent image. Since the cameras have a conical field of view (see FIG. 7, reference number 29 ), from a certain height 28 there will always be an overlap of the image fields, which the image evaluation unit 10 must also take into account.

The calculated coordinates of the process units located in the area of an evaluation unit 3 , 7 are provided by the latter with the aid of a network connection 11 for further use on a network 4 . This further use can be, for example, a control process or quality data acquisition. The network 4 can, in particular, also serve to logically link several camera arrays 1 , 5 with their associated evaluation units 3 , 7 and thus quasi form a macro array or super array from individual arrays. Logical calibration information, for example the crossing of a process unit over a camera array boundary into an adjacent array, flows between the evaluation units involved.

Furthermore, a central control unit can also be used zel arrays coordinate, control and the movement data manage. This corresponds to a control of the macro  Arrays or super arrays by a central control unit unit or superordinate intelligence.

In the following the operation of the device shown Vorrich is explained in more detail with reference to FIGS . 4, 5 and 6.

The cameras of the camera array 1 , 5 are preferably mounted on the ceiling of a production facility and record the scenery from above. In FIG. 4, six cameras fifth total arranged in a grid on the monitor the evaluation region. The scenario shown also includes a lock 13 through which process units are introduced into the clean room area that is present in the production facility.

The already known detection options such as transponders 12 , 16 , barcode, infrared identification or other identification of the production units or the virtual data transfer based solely on information processing (databases etc.) transfers to the optical system according to the Invention which belongs to the process unit is the data. This takes place either at the lock 13 or at other defined identification points 15. This enables a clear identification at the starting point of a space trace, so that process identifications or data can be clearly assigned to a process unit moving in a space trace.

Once this assignment has been made, the operator 18 can remove the process unit from the lock and place it on a table 14, for example. This is the simplest example of a chaotic interim storage facility. Re gale or more complicated facilities can be detected with a three-dimensional evaluation. The process unit is moved on a space track 19 , which is shown in FIG. 5. The system now has four process units in the detection area, the associated data of which can also be managed by the system.

According to FIG. 6, the operator 18 now takes a different process unit 17 back from the table 14 and moves it to the space track 21 to input station 15 of a location Prozeßan. At point 22 there is the logical continuation of the space track with which the process unit was conveyed to the table 14 .

At the defined end of the space track 21 , the assignment back or a check of the assigned data and possibly a data transfer to a subsequent or superordinate process environment takes place. If the transponder identification 16 is different at the beginning and at the end of a space lane without a change in the transport media (transponder) being provided, then an error situation has obviously occurred.

Since the system constantly saves or collates oneline where a process unit is located, it is also possible to define logical restrictions on the movement. This is done by defining movement corridors, these movement corridors being characterized by data stored in a memory of the image evaluation unit 10 . If a process unit is moved beyond such a virtual limit, the system issues an alarm signal. This prevents a process unit from being accidentally removed from the process area and placed somewhere and / or forgotten.

Acquiring a three-dimensional space curve is more difficult than acquiring a two-dimensional space curve. The third spatial coordinate must be extracted from the overlap of two camera images, as illustrated in FIG. 7. Each evaluation of a camera channel provides a two-dimensional set of coordinates with the position of the process unit in the image. What cannot be easily extracted from the individual images is the distance of the object 26 from the camera, ie the height position of the object 26 in space. If there is only a two-dimensional image, then the height position of the object can only be determined by the detected object size, which is very expensive.

The height position of the object in space can be determined more easily if one carries out geometric evaluations, taking into account the arrangement of two cameras 23 and 24 , evaluating their images or the already extracted 2D coordinates (preprocessing). For example, the viewing angles α and β of the two cameras 23 and 24 can be calculated and the height (third spatial coordinate) determined from a simple triangulation.

The prerequisite for this is that the object 26 is detected by at least two cameras in each position in the area to be monitored. For this purpose, it can be advantageous to install at least one camera also tilted and / or rotated, as is shown with reference to the camera 25 in FIG. 7. This means that the geometric effort for the evaluation is more complex, but the fields of view of the cameras can be used better and more effectively. However, the prerequisite is that the patterns to be carried out can analyze and evaluate all perspectives of a process unit.

If you work with an optical system, it is also possible to capture and read defined optical markings. If the process units 30 , as shown in FIG. 8, are additionally provided with a bar code 32 , a 2D code (bitmap) or an OCR font 31 , their detection, evaluation or reading can also be carried out within the scope of the Motion detection according to the invention take place.

This has the advantage that the presence and identification of a process unit 34, for example at a receiving point in this equipment, can be detected in process machines or process systems (equipment) and - if there is a network - can also be forwarded to the equipment. This is particularly advantageous if there are difficulties in the vicinity of the receiving point 35 to attach antennas or other attachments for identification, which restrict the freedom of movement around the receiving point 35 . A camera over the equipment does not interfere. The easiest way - as already explained above - is to install the cameras on the ceiling of the respective room.

Of course it is also possible to use a camera umpteen for the purpose of presence and identification use acquisition without using space at the same time analyze traces.

Claims (27)

1. A method for detecting the movements of Prozessein units during a production process in a given evaluation area, characterized in that by means of at least two fixedly arranged in the evaluation area, a camera array forming cameras and an evaluation unit, location coordinates of each process unit are continuously detected that a coordinate describing the movement of the respective process unit is determined for each location coordinate, and that corresponding data for each process unit is determined by stringing together the determined translation vectors of a spatial trace. 2. The method according to claim 1, characterized records that the data corresponding to the space track can be saved. 3. The method according to claim 1 or 2, characterized ge indicates that a two-dimensional space trace corresponding data can be determined. 4. The method according to claim 1 or 2, characterized ge indicates that a three-dimensional space trace corresponding data can be determined. 5. The method according to any one of the preceding claims, characterized in that in the evaluation is the data provided by the cameras be compared to a sample. 6. The method according to claim 5, characterized records that the cameras always the process units grasp from the same perspective.   7. The method according to claim 5, characterized characterizes that the cameras distinguish the process units different perspectives. 8. The method according to any one of the preceding claims, characterized in that the output signals each camera is supplied with its own evaluation unit the. 9. The method according to any one of the preceding claims, characterized in that the output signals the evaluation unit to a further evaluation unit are led, with the further evaluation unit to the Er summarize the movements of the process units in a wide range evaluation area with the cameras of another camera mera arrays is connected. 10. The method according to any one of the preceding claims, characterized in that when a Process unit not previously registered a message signal is produced. 11. The method according to any one of the preceding claims, characterized in that the of several Evaluation units provided output signals are fed to a central control unit. 12. The method according to any one of the preceding claims, characterized in that at an input and / or the starting point of the specified evaluation range Transfer data to identify a process unit for Will be provided. 13. The method according to any one of the preceding claims, characterized in that movement corridors corresponding data are specified and an alarm signal is generated when a process unit the specified Movement corridor leaves.   14. The method according to any one of the preceding claims, characterized in that by means of the cameras and the evaluation unit also continue to process units applied process unit identifiers the. 15. Device for detecting the movements of pro cess units during a production process in a predetermined evaluation range, with

• - At least two stationary in the evaluation area arranged th, a camera array forming cameras, and
• - An input side connected to the cameras from the evaluation unit, which continuously detects location coordinates of each process unit, determines a movement vector describing the movement of the respective process unit for each location coordinate, and determines corresponding data for each process unit by stringing together the determined translation vectors of a space trace.

16. The apparatus according to claim 15, characterized indicates that the specified evaluation range Is clean room area and the cameras on the ceiling of the Clean room area are arranged. 17. The apparatus of claim 15 or 16, characterized characterized that the cameras are analog cameras. 18. The apparatus of claim 15 or 16, characterized characterized that the cameras are digital cameras are. 19. Device according to one of the preceding claims 15 to 18, characterized in that the Ka meras connected to the evaluation unit via a data bus they are.   20. Device according to one of the preceding claims 15 to 19, characterized in that the evaluation unit has a series circuit which consists of a multiplexer ( 8 ), a digitizing unit ( 9 ) and an image evaluation unit ( 10 ). 21. The apparatus according to claim 20, characterized in that the image evaluation unit ( 10 ) is a digital signal processor (DSP) or an FPGA (Field Programmable Field Array) or has both a digital signal processor and FPGA. 22. Device according to one of the preceding claims 15 to 21, characterized in that everyone Camera is assigned its own evaluation unit and the Evaluation units connected to one another on the output side are. 23. Device according to one of the preceding claims 15 to 22, characterized in that they meh has more camera arrays and several evaluation units and that at least two of these evaluation units are included are connected. 24. Device according to one of the preceding claims 15 to 23, characterized in that they have a Central control unit that controls all Licher evaluation units is provided. 25. Device according to one of the preceding claims 15 to 24, characterized in that the Image evaluation unit for capturing processes provided applied process unit identifiers is.   26. Device according to one of the preceding claims 15 to 25, characterized in that the Image evaluation unit a memory in which a loading corresponding corridor corridor data is stored, and alarm means which emit an alarm signal ben when a process unit the specified movement leaves corridor. 27. Device according to one of the preceding claims 15 to 26, characterized in that mind at least one of the cameras is tilted or twisted is.