Joint International IMEKO TC1+ TC7 Symposium September 21− 24, 2005, Ilmenau, Germany

DYNAMIC VISION TRAINING MODULE FOR i-LEARNING IN

INDUSTRIAL APPLICATIONS

Uwe Nehse1, Franziska Annemüller1 and Dietmar Ernst2

1Mahr OKM GmbH, Carl-Zeiss-Promenade 10, D-07745 Jena, Germany, info-okm@mahr.de

2esco , Jülicher Strasse 134, D-52070 Aachen, Germany, info@esco-aachen.de

Abstract Vision Training modules are used in the measurement

of complex components and tools (e.g., milling cutters, drills) with precision coordinate measuring machines using optical image-processing sensors in closed-loop manufac-turing. The closed loop comprises the machine-tool, the measuring machine and the software packages belonging to them.

Few scientific studies exist in the field of Vision Train-ing modules for i-Learning. For the kind of machines or instruments mentioned above it is common to supply opera-tor instruction manuals in electronic form. These instruc-tions have no modular structure and are not related to influ-ences from the environment of the machines/instruments, their operators, and the manufacturing jobs. Training is by way of rather general, text-based instructions. This hardly enables users to adapt jobs on their own or to solve new tasks with the existing system.

The paper presents innovative operating instructions and training modules for i-Learning, which include dy-namic components. As a basis for the dynamic approach, a concept has been developed that imparts a modular struc-ture to the existing operating and training instructions. The influences taken into account are the configuration of the hardware and software, the objects under test, the applica-tion, and the state of knowledge of the operators.

The application of the dynamic Vision Training mod-ules is explained with the example of the optical coordinate measuring machines made by Mahr OKM GmbH and the HAWK/OSPREY software components that go with them, as used in the closed-loop machining of tapered precision valves.

Keyword

i-learning, vision training module

1. INTRODUCTION Manufacturers of precision tools increasingly re-

quire that measuring equipment be integrated in their existing manufacturing line. As a result, users at different levels differ in their basic knowledge about

facts and relationships in metrology and in quality management. Therefore, dynamic training concepts are required, based on Vision Training Modules for i-learning systems [1]. In this, instrument hardware and software configuration, the measurement job and the operators' state of knowledge are decisive influence factors.

Training materials are prepared in a dynamic mode, for different categories of measuring machine users having different levels of knowledge. This should enable every category of users to perform the respective measuring jobs, make appropriate adapta-tions of the procedure, and use existing systems for solving new measuring tasks.

It is intended, in the near future, to use such Vi-sion Training Modules for i-learning systems in the measurement of precision tools such as drills or mill-ing cutters, or other precision components (e.g., rota-tionally symmetric turned parts) by means of coordi-nate measuring machines integrated in the process quality control loop. As far as the manufacturing of tools with complex geometries (e.g., hob cutters) is concerned, a differentiation must be made between large and small quality control loops.

The small quality control loop is used for the permanent control in the manufacturing process by direct influence on the various machining steps (ma-chine-tool, machining cell, machining process). This type of control loop has a direct, in-process influence on the quality characteristics to be established [2].

The large quality control loop is intended for post-process checking and quality verification by a delayed influence on several manufacturing stages (machining, assembly, final inspection and service). These measures have an indirect influence on the quality characteristics to be established, i.e. an influ-ence on the quality of future work. Within limits, they allow corrective action to assure the quality of products already made [2].

FEDiii

...ZA...EIMEKOBHOME

Quality control loops are needed to implement quality management requirements, e.g., to control and continually improve the processes with regard to product quality. Fig. 1 shows the two closed quality control loops with their basic functions. Also shown are some other important functions such as display of measured data, statistical analysis, and documentation of measurement and inspection in the process cell. The individual functions are linked by paths; several paths may be active in parallel. The segment mount-ing – work (testpiece) – measuring element is com-mon to the entire process.

Fig. 1: Basic functions in testing and feedback [3]

In order to implement a controlling element in the

process quality control loop comprising a machine-tool and a measuring instrument, the latter must be relocated from the inspection room to the manufac-turing line (Fig. 2).

Fig. 2: Process quality control loop as, for example, in the manufacture of cutting tools

This closed quality control loop (Fig. 3) has the

following structure. The controlled system is repre-sented by the machine-tool. The specified quality characteristics of machined parts are sensed by the quality measuring element; here, this is a Mahr OKM GmbH multisensor coordinate measuring machine with OSPREY measurement software module. The quality controller is the HAWK software module

supplied by esco GmbH. This module is intended

for the digital measurement of geometrical quantities as an information base, a controlling element in the machining process control loop, and as a link be-tween CAD systems, machine-tool and measuring instrument, permitting superordinate information processing. The quality correcting element depends on the comparison between specified and actual sizes; in the present process, it is an NC wheel dressing program.

Fig. 3: Block diagram of a process-integrated quality control loop

2. STATE OF THE ART IN COORDINATE

MEASUREMENT Fig. 4 shows a user-specific, process-integrated

closed control loop used in the machining of preci-sion tools.

The process cell is mainly used for the generation of new tool geometries, for tool resharpening (e.g., the redressing of grinding wheels), or for controlling the process taking place in the manufacturing line. By means of such a process cell, information obtained through the quality measuring element, the quality controller and the quality correcting element is fed back to the machining process. Here, the cell is di-vided into three operational stages.

Fig. 4: User-specific, process-integrated closed control loop

FEDiii

...ZA...EIMEKOBHOME

Operational stage 1 comprises offline inspection planning / machining planning. For its successful use, various interactions with the user or with other systems are necessary. For the implementation of a user-specific measurement of the tools machined, certain prerequisites have to be established by the user in advance. These include inspection planning by means of CAD-based analysis software or specific measurement programs for standard tools. It is also possible for the user to prepare his inspection sched-ule offline, which means that the function of the ex-isting process cell is not affected.

Operational stage 2 is the required integration of measuring equipment into a manufacturing quality control loop. This requires that the measuring equip-ment features

� immunity to environmental influences (encapsulation, vibration damping), � easy operation, � high availability, � short measuring times, � low measuring uncertainty, and � data networking facilities.

Thanks to the many different sensor systems

available (optical as well as tactile), coordinate meas-uring machines provide an efficient basis for the reliable detection of all specified quality characteris-tics.

Operational stage 3 comprises the online / offline visualization of the quality controller. Here, the qual-ity characteristics (see Fig. 1) are documented and analyzed with varied visualization facilities. By means of diverse statistics software modules, quality characteristics of the measuring instrument, such as the gauge potential index cg, gauge capability index cgk or the gauge repeat and reproducibility index GRR, can be analyzed statistically.

3. STATE OF THE ART IN TRAINING AND

DOCUMENTATION 3.1. Modular documentation of coordinate meas-uring machines Multisensor coordinate measuring machines are of

modular design. Thanks to this special property, the configuration of their hardware components and the software modules is variable. This means that certain hardware components such as the optical measuring head with zoom lens or interchangeable lenses, or the horizontal rotary axis unit (Fig. 5, Fig. 6) can be used with the various machine models in a modular fash-ion.

Thanks to their modularity, coordinate measuring machines are suitable for shop-floor use and can be easily configured by the user for the measuring job in hand and, so that near-process measurements can be made with the required accuracy. The main field of

near-process application of these machines is the measurement of precision cutting tools (hobs, ball track milling cutters), rotationally symmetric parts or other precision components.

It is evident that the modularity of these measur-

ing machines supports a modular structure of dy-namic training instructions.

3.2. Documentation and training in coordinate

measurement Current training materials in coordinate measure-

ment differ in their structure from manufacturer to manufacturer, and they assume a high level of previ-ous knowledge about metrology and coordinate measurement. As a rule, training is imparted by text-based training materials, which are available in dif-ferent Office formats, such as MS Word or MS PowerPoint. This prevents their mutual combination; in other words, the texts are not dynamic.

Dynamic training materials enable the user to reach the training goal faster by means of hyperlinks or by linking the materials, without having to strictly follow the given training concept. The dynamic prop-erty can just as well be described as a modular struc-ture of the training materials, which allows training modules to be exchanged.

The requirement for comparable and reproducible measurement results with low uncertainties, and the need for continuous quality improvement and quality cost reduction make user-specific training a must, considering that the operators in a process cell have different levels of knowledge about measurement techniques, quality management and design.

To promote comprehensive, sound, comparable and documented training in coordinate measurement that satisfies current needs irrespective of equipment manufacturers, the AUKOM e.V. association was founded. Its members are well-known manufacturers and users of coordinate measuring machines.

The users of coordinate measuring machines to-day are required to have a high level of knowledge about metrology, design, quality assurance, meas-urement software and measuring machines (Fig. 7).

Fig. 5: UNI-VIS with horizontal rotary axis unit [4]

Fig. 6: UNI-VIS without horizontal rotary axis unit [4]

FEDiii

...ZA...EIMEKOBHOME

Fig. 7: Operator knowledge required [5] To minimize the deficiencies of currently avail-

able training, AUKOM e.V. has developed a product-neutral, standardized concept of process cell training, covering several stages [5].

In AUKOM – Stage 1, operators of coordinate measuring machines (CMMs) are imparted basic knowledge in metrology, design and quality man-agement. After this stage, the trainee is able to read technical drawings, distinguish between different types of CMM, qualitatively estimate uncertainties of measurement due to influences from the work, the environment, the operator and the measurement strat-egy, and thus to minimize possible sources of error [6]. Based on this, AUKOM Stage 2 is a user training module. Here, the trainees learn how to derive meas-urement jobs from technical drawings, to define ap-propriate measurement and probing strategies, to estimate potential measurement uncertainties in quan-titative terms, and to reduce them. Other important subjects at this stage are the writing and documenta-tion of CNC measurement programs, and the applica-tion of the principles of quality management [7]. The highest level, AUKOM Stage 3, assumes trainees who have completed stages 1 and 2 and trains them to become CMM experts. CMM experts must be able to proficiently verify measurements and measurement results, apply a broader range of tools in quality man-agement, etc [8].

At present, manufacturers of CMMs are starting to implement a visualized solution of measurement jobs for their customers. Shown below is a section from an existing training material entitled „Measure-ment of profiled parts“ [4]. The section headed „De-fining the probing direction in HAWK“ (Fig. 8) illus-trates the visualization implemented in the use of the HAWK software module in connection with a particu-lar measuring job. The visualization provides the possibility to point to special settings in the hardware or software.

Fig. 8: Section from the training material „Measurement of profiled parts“ [9]

4. WHY A NEW TRAINING CONCEPT? As a consequence of the relocation of measure-

ment to the manufacturing line, the coordinate meas-uring machines (CMMs) are operated by line or ma-chine-tool operators. These are people responsible for setting up, operating and converting the various ma-chines (machine-tools, measuring machines) in the process cell.

These operators have little know-how about me-trology and quality management. Therefore, the exist-ing general training materials provided by the various CMM manufacturers are not suited to this group of users.

The integration of measuring equipment into the manufacturing line creates a process cell. Within such a cell, the flow of materials and data has to be divided into a number of process steps. According to these process steps, suitable training steps have to be de-veloped. In other words, modular training concepts are required for different user groups and topics, to enable these users to do their job efficiently and to continuously improve product quality.

5. DESCRIPTION OF THE NEW CONCEPT To develop a modular training concept for this

particular process cell, it was necessary to make a general distinction between the different user groups and topics.

The user groups are distinguished by such criteria as previous training, present level of knowledge in metrology, quality management, design and informa-tion technology. A highly important criterion is the scope of responsibility within the process cell. Ac-cording to these criteria, the following groups are distinguished:

� Operators � Programmers in manufacturing

(shift managers, machine-setters)

FEDiii

...ZA...EIMEKOBHOME

� Programmers in quality management (shift managers, machine-setters)

� System supervisors (department managers)

Most tool manufacturers require CMMs in a proc-

ess cell to feature single-button operation or menu-driven operation.

To ensure general and specific training in the principles of coordinate measurement, quality man-agement and certain areas of design, the AUKOM stages described should go into this training concept. Section 6 will present a concept of a dynamic training module for different user groups, exemplified by a near-process measurement job.

As the topics in optical precision measurement are many and varied, these topics should be distinguished according to a number of criteria as well. These are

� CMM hardware configuration � OSPREY measurement software � HAWK analysis and documentation

tool � Measurement job � Level of knowledge of user groups

To illustrate the topic CMM hardware configura-

tion, we have selected the calibration of sensor sys-tems (Fig. 9). A distinction is made between the cali-bration of optical and that of tactile measuring sys-tems. For the calibration of optical measuring heads, zoom lenses have to be distinguished from lenses with fixed focal lengths. Both during installation and during training, exactly those modules have to be combined that are required by the user and the jobs intended. The user should master not less and not more than those hardware modules that are actually installed on his CMM.

Fig. 9: Concept exemplified by the calibration of sensor systems

To be able to develop this new type of dynamic

training modules for i-learning systems, it is impera-tive to make use of modern information processing technologies.

The aim should be to prepare the training materi-als in uniform formats. The formats should be stan-dardized to allow their fast and easy communication via the WorldWideWeb. Thanks to the improved availability of the Internet, i-learning systems are becoming increasingly interesting for the user groups

within the process cell. Whether i-learning is defined as interactive learning or individual learning, it is a flexible way of electronic learning [10].

For a CMM manufacturer to impart knowledge and training by way of i-learning systems, means that the training materials have to be continuously up to date. Modular training with i-learning systems means the simultaneous training of participants actually at different locations but virtually in one and the same room [11].

One can imagine that, in the near future, suppliers of accessory modules to CMMs or of software pack-ages will supply their part of the operating and train-ing instructions covering their product.

6. CONCEPT FOR A DYNAMIC TRAINING MODULE FOR DIFFERENT USER GROUPS This section describes the modular structure of

training materials for different user groups, exempli-fied by a shop-floor measurement job.

The shop-floor measurement job is a typical in-process measurement occurring during the machining of hob cutters within a process quality control loop (Fig. 10). The measurement centre within the process cell comprises a coordinate measuring machine made by Mahr OKM and the associated OSPREY image processing software. Within the quality control loop, the HAWK programming, data analysis and docu-mentation software module, a product of esco, con-stitutes the quality controller (comparison, and gen-eration of the variables to be controlled).

Fig. 10: Process control loop Machining of hob cutters In view of user groups with different levels of re-

sponsibility, a number of separate dynamic training modules have been created, which form a system of interlinked stages.

The modules are organized as described in section 5.

Each training module is a combination of an AU-KOM training stage with a machine-specific training module of Mahr OKM. In the following, the training modules for the user groups will be described.

FEDiii

...ZA...EIMEKOBHOME

6.1 Training module for the group of Operators Before a dynamic training module for this group

can be created, it is necessary to analyze the user-specific requirements.

Within their process control loop, tool manufac-turers demand that the measuring instrument or ma-chine features single-button or menu-driven opera-tion. For this group of users it is assumed that their existing knowledge of metrology, quality assurance and design is low. Through the AUKOM Stage 1 training module, this group is taught the principles of quality assurance (e.g., how to avoid potential error sources) and of metrology.

Training in how to do the basic setting of the OS-PREY and HAWK software modules must be given going by machine-specific instructions. As far as OSPREY is concerned, major emphasis is placed on light control of the measuring machine. The training in HAWK must include such aspects as how to open a measurement job, how to start a measurement and, finally how to release corrective action.

A training subject that is just as important is how to respond to error messages.

The training module for this user group is sche-matically illustrated below (Fig. 11).

Fig. 11: Schematic diagram of the training module for the group of Operators proper

6.2 Training module for the group of Program-

mers in manufacturing and quality manage-ment

The main task of this group within an existing process cell is the development of different measure-ment strategies for the multisensor systems used. These people need a fundamental understanding of the mode of operation of edge locus and focus crite-ria, an understanding of design data, as well as know-how in quality management. The Programmer mod-ule assumes successful completion of the Operator

module. The theoretical aspect, irrespective of the make of the measuring equipment, is trained by means of the AUKOM Stage 2, whereas practical training must include special submodules of the measuring machine manufacturers and software pro-viders that teach how to calibrate the respective hardware and software components.

The modular training concept for this group is shown schematically in Fig. 12. Main emphasis in this module is placed on settings to be made in the OSPREY measurement software, calibrating instruc-tions for the hardware components used, and training in how to use the HAWK software. It is important to ensure that this training module is explicitly related to the machining process; this means the requirement to assign and parameterize the various grinding wheels and wheel dressing tools employed.

Fig. 12: Schematic diagram of the training module for the Programmer group

6.3 Training module for the group of

System Supervisors This is the top level of users. These people need to

understand everything about the process cell, the interaction between measurement technique and ma-chining, and quality management. This is taught through AUKOM Stage 3 in connection with specific training materials be the measuring equipment and software makers. As the structure illustrated in Fig. 13 shows, the training module for this group assumes successful completion of the two previous modules, i.e., those for the Operator and Programmer groups.

Another subject of prime importance in this mod-ule is the learning of the directory tree structures of

FEDiii

...ZA...EIMEKOBHOME

the software systems employed, and the correct han-dling of the settings to be made in HAWK for the respective measurement techniques and procedures.

All in all, the user at this level must have a com-plete overview and knowledge of the hardware and software concepts of the measuring machines or in-struments as well as of the machine-tool used in the process cell.

Fig. 13: Schematic diagram of the training module for the group of System Supervisors

7. SUMMARY The intention of the modular training concept de-

scribed is to match learning processes to the current qualification requirements of individual levels of machine operators. The training measures are con-ceived as on-the-job training in combination with Rapid e-Learning.

With regard to information technology, the vari-ous training modules are generated in standardized formats. Therefore they can be readily communicated via the Internet. The method has five obvious advan-tages:

1. Training is getting individualized. 2. Training is independent of place and ti-

me. 3. The training modules can be updated at

short notice. 4. Training costs incurred by the manufac-

turers and users of measuring equipment are considerably lower compared to clas-sical training methods.

5. Training contents are exactly matched to the user's hardware modules.

The new concept of learning makes it possible to

hold virtual training courses that permit a synchro-nous, simultaneous training of several participants at different locations.

In summary, i-learning systems are the further education medium of tomorrow.

REFERENCES [1] Weckenmann, A.; Hoffmann, J.: Quality in

Metrology - Metrology for Quality. IMEKO 8th International Symposium on Measurement and Quality Control in Production ISMQC (11.-15.10.2004, Erlangen), In: VDI-Berichte Nr. 1860, Düsseldorf: VDI-Verlag, 2004.

[2] Linss, G.: Qualitätsmanagement für Ingenieure. 2. Auflage Leipzig: Carl Hanser Verlag, 2005.

[3] Masing, W.: Handbuch der Qualitätssicherung. München Wien:Carl Hanser Verlag, 1980.

[4] Mahr OKM: MarVision Optische Koordinaten-messtechnik. Produktkatalog, Carl-Zeiss-Promenade 10, 07745 Jena, Germany: Mahr OKM GmbH, 2005

[5] AUKOM - Ausbildung Koordinatenmesstechnik e.V. http://www.aukom.info/deutsch/wissen.htm 2004 [6] AUKOM - Ausbildung Koordinatenmesstechnik e.V. http://www.aukom.info/deutsch/stufe_1.htm [7] AUKOM - Ausbildung Koordinatenmesstechnik e.V. http://www.aukom.info/deutsch/stufe_2.htm [8] AUKOM - Ausbildung Koordinatenmesstechnik e.V. http://www.aukom.info/deutsch/stufe_3.htm [9] OKM: Messen von Profilteilen. Trainingsunterlage,

Carl-Zeiss-Promenade 10, 07745 Jena, Germany: OKM GmbH, 2004

[10] iLearning – Gesellschaft für innovatives Lernen mbH. http://www.ilearning-company.de . 2004

[11] Mayer, T.:i-Learning statt e-Learning. Dissertation Uni-Erlangen. 2001

http://www.opus.ub.uni-erlangen.de/opus/volltext/2004/13 2004

[12] Taylor, J.R.: An introduction to error analysis, University Science Books, Sausalito, CA, 1997.

[13] Hofmann, D.: Handbuch Messtechnik und Qualitätssicherung. Berlin: 3. Auflage Berlin: Verlag Technik, 1986.

Author(s): Dipl.-Ing. (FH) F. Annemüller (Mahr OKM GmbH, Carl-Zeiss-Promenade 10, D-07745 Jena, Germany), franziska.annemueller@mahr.de Dr.-Ing. U. Nehse (Mahr OKM GmbH, Carl-Zeiss-Prome-nade 10, D-07745 Jena, Germany) uwe.nehse@mahr.de Dipl.-Ing. D. Ernst (esco GmbH, Jülicher Strasse 134, D-52070 Aachen, Germany) dietmar.ernst@esco-aachen.de

FEDiii

...ZA...EIMEKOBHOME