Since the late 60s image processing systems have entered the domain of industrial applicability. In the past starting with poor performance, today they reach for a high level of usability and standard. If it comes to rapid prototyping applications measurement systems are looked for that allow for the precise, comprehensive, flexible and hopefully standardized use of measured data. Image processing integration gives rise to new systems that meet these requirements. For two-dimensional measurement tasks it is easy to find a standardized interpretation of the measured data. Standards for these geometric considerations relate to the achievements of coordinate measurement systems with tactile sensors. Lines, intersections, circles, ellipses are easy to calculate, even with a statistical interpretation in mind. Much more difficult is the interpretation of three- dimensional data. This paper deals with the standardization efforts related to the interpretation and evaluation of data acquired by image processing based measurement systems. Systems capable of measuring 'point clouds' in the three- dimensional space. Transferring them into representations useful for machining procedures.

Imaging measurement techniques are preferably applied to digitizing and measuring of complex three-dimensional scenes. Topometric techniques allow imaging 3D-data acquisition within a few video cycles or even in video real- time. Integration of the topometric systems into coordinate measuring- and handling machines is supported by compact and light sensors. These sensors can be optimized for a specific task with respect to accuracy, field size and operating distance. An adaptive data acquisition can be achieved by using a multi-field sensor. In the field of surface inspection topometric systems are used for quality control within a production line.

In the industrial sector dimensional measurement within scenes of the order of one square meter is required. These measurements may be achieved with active structured lighting. Programmable optical grids can be projected directly on the surface of unpolished test-pieces in order to realize an absolute measurement method called 'coded light approach' to obtain a full depth-image, requiring only a small number of projections. In combination with the phase shifting method subpixel accuracy is achieved. Geometry- checking and controlling the surface quality of objects with mirror-like surfaces is put into practice with indirect illumination. Using a line-projector, stripes are projected to a diffusely reflecting surface, e.g. a wall. The test piece is positioned in a way that the lines are seen indirectly by the camera, reflected in the surface.

Computer aided technologies in industry increase the demand for 3D digital data. To support industry, photogrammetry offers a wide field of methods, from the classical one based on stereo photogrammetry up to raster projection and Moire methods. The paper describes an entirely digital solution for gathering 3D coordinates of surfaces and its use for rapid prototyping. This method uses a combination of digital cameras, interactive and automatic measurement procedures, bundle adjustment and digital image correlation to produce a 3D surface model. For recurring applications an automatic run through all parts of the software is possible, which is a prerequisite for its use by workers in production. The car and steel industries are two typical fields for this type of application.

Because of its process quality and high flexibility, the high power laser has become established in many different fields of material processing. In order to reduce the time required for the programming of 3D laser systems computer based offline-programming systems are necessary. An offline- programming module is being developed to enable a swift generation of high processing quality programs for 3D-laser systems. Offline-programming is possible for gantry systems and industrial robots, with robots being more and more important for 3D laser processing, because of their lower investment cost. In order to achieve a high process quality, the aim is to generate technologically adapted NC/RC- programs. With offline-programming for robots, it is significant to consider the robots cinematic properties already in the process of program generation. As the motion behavior and the accessibility significantly depend on the position of the workpiece, it is an important task to place the workpiece appropriate. For that purpose a layout optimization module takes into account the properties of the machine and the influence of the process itself. The offline-programming module is based on a 3D-graphical simulation system, so that by means of realistic simulation, the real production situation can be considered.

Concerning laser welding as a manufacturing process, there are still extensive experimental tests necessary to determine the resulting product properties in order to optimize the manufacturing process and to improve the product design. A simulation of the laser welding process can contribute to supply information about achievable product properties already in the very early stages of product design and production planning. As a result, considerably more reliable planning results can be obtained, while at the same time the necessary experimental expense decreases. The aim of process simulation, based on finite element methods, is to gain information about properties relevant to product design, e.g. the geometry of the seam cross section or the distortion of the workpiece. Due to the extremely large temperature gradients within a small volume of energy absorption, it is necessary to analyze the process in two steps. In the first step, the welding process is modeled in a very detailed way within a heat transfer analysis. Already the resulting time-temperature behavior of the material provides information about microscopic properties of the weld seam, e.g. cross section geometry or hardness. Based on those results, the heat influence on the whole workpiece, e.g. residual stress and distortion, can be determined in the following second step, using a coupled thermo-mechanical analysis with a less detailed model. The results of the analysis coincide well with experimental data.

Three-dimensional-measurement of objects is an integral part in rapid prototyping and reverse engineering. In the last years, topometric sensors have gained increasing usage in this field. The provide an optical method for a fast and robust calculation of dense 3D-point clouds. This paper presents two concepts for the 3D-measurement of complex objects. They offer functional solutions for digitization of large and complex objects with high accuracy, but with low mechanical and financial effort.

Optical 3-D measurement is an interesting approach for rapid prototyping. On one hand it's necessary to get the 3-D data of an object and on the other hand it's necessary to check the manufactured object (quality checking). Optical 3-D measurement can realize both. Classical 3-D measurement procedures based on photogrammetry cause systematic errors at strongly curved surfaces or steps in surfaces. One possibility to reduce these errors is to calculate the 3-D coordinates from several successively taken images. Thus it's possible to get higher spatial resolution and to reduce the systematic errors at 'problem surfaces.' Another possibility is to process the measurement values by neural networks. A modified associative memory smoothes and corrects the calculated 3-D coordinates using a-priori knowledge about the measurement object.

Existing systems for computer integrated manufacturing are based on the principle of the process chain: The product runs through different production sections as design, work planning and manufacturing in a sequential order. The data generated by a production sequence are transferred via interface to the following production sequence. These tightly-packed production sequences leave little scope for responding to quality deviations. This deficit is highlighted particularly in 3D laser cutting processes. In order to achieve an optimum machining result, a series of preliminary tests is required. Quality control loops play an important role in restricting the scope of necessary testing to a minimum. The represented control loop contains a CAD- system to design the workpiece, an offline-programming system to develop working strategies and NC/RC-programs as well as a shop-floor oriented tool to record quality data of the workpiece. The systems are coupled by an integrated product model. The control loop feeds quality data back to Operations Planning in the form of rules for processing strategies and technological data, so that the quality of the production process is enhanced. It is intended to supply optimum process parameters, so that the number of preliminary tests can be reduced. On the other hand the control loop contributes quality enhancement measures which serve as rules for the designers.

In this paper, we describe a new automated inspection method of parts produced by rapid prototyping machines using a 3-D range sensor. The input to the program is a tessellated representation of the part at a desired resolution saved in the standard STL format and an un-ordered series of measurements produced by a 3-D optical sensor. The output is an inspection report indicating the level of discrepancy between the measured points and the model. Using this inspection system an operator of a rapid prototyping machine can rapidly identify fabrication defects or monitor process drift during conversion process. At the base of the method, a new robust correspondence algorithm which can find the rigid transformation between the tessellated model of the part and the measured points, is presented. This method is based on a least median square norm capable of a robustness of up to 50%. The robustness of the method is essential since one cannot guaranty that in practice, all the points in the measured set belong to the model. These types of algorithms are usually quite costly in computational complexity but we show that one can speed-up these algorithms by using the well-known iterative closest point algorithm (ICP) and a multi-resolution representation scheme based on voxels. We analyze in detail the complexity of the algorithm. We also present experimental results on a complex part with global and local inspection processes.

The application of an active vision 3D sensor is described for the development and control of an autonomous intelligent robot cell for the disassembling of end-of-life-vehicle components. The research and development work was done concurrently by three European development teams at different locations. During this phase, the virtual environment was distributed on the local development platforms of these teams. Intermediate development results and 3D sensor data were exchanged through network communication to be mutually tested and verified. The physical environment of the disassembling cell demonstrator and its sensor systems is currently being integrated at the BIBA institute.

Fast technological advances and steadily increasing severe worldwide competition force industry to respond all the time faster to new and chanced customer wishes. Some of the recently emerged processes, commonly referred to as 'rapid prototyping' (RP), have proved to be powerful tools for accelerating product and process development. Early approaches aimed at the automated production of plastic models. These techniques achieved industrial maturity extremely fast and are meanwhile established as standard utilities in the field of development/design processes. So far, their applicability to metal working industry was limited to design studies because the mechanical properties of the prototypes, e.g. modulus of elasticity and mechanical strength were not comparable to the final products they represented. Therefore, RP-processes aimed at the direct production of metallic prototypes gained more and more importance during recent years. A technique belonging to this group is manufacturing of prototypes by using a laser beam sintering machine capable of directly processing metal powders. This so called laser beam sintering process showed a great potential for direct manufacturing of functional tools and prototypes in early feasibility studies. Detailed examinations were performed at several research centers to determine the attainable quality of the parts concerning roughness, dimensional accuracy and mechanical strength. These examinations showed, that there still is a considerable demand for quality improvements of the previously mentioned parameters. The practical application and the potential for improvement of the geometrical accuracy of laser beam sintered parts by using a dual beam concept was proven. An innovative beam guiding and forming concept, similar to the previously mentioned patented beam guiding system, was developed and built with the goal to improve the process parameters governing mechanical properties as well as geometrical accuracy. Further reaching technical applicability of this concept for various other applications is certainly conceivable.

Laser based rapid tooling techniques enable the completion of tools within a minimized period of time. Especially the controlled laser ablation process of metals or ceramics allows precise manufacturing along with a high surface accuracy of the parts. The reactive ablation mechanism of ferrous materials in oxygen atmosphere -- the chip removal - is described, as well as the optimization of the process parameters by systematical procedures. The processing results are limited with given radius of the interaction zone between the laser beam and the workpiece surface. Essential process-parameter is therefore the focus radius. This parameter, effectively being controlled by an adaptive optical mirror, strongly influences the process and thus the workpiece result. The laser ablation process characteristically offers a high flexibility concerning workpiece materials and geometries. In combination with the ability of processing even hardened steels without any tool wear laser ablation is predestined for the rapid tooling of metal forming tools like hot forging dies. Obtaining high workpiece accuracy along with short manufacturing times recommends the optimized ablation process not only for prototyping but for serial manufacturing as well.

Mobile systems in the area of modern manufacturing become more and more important. In order to handle an object with a manipulator mounted on an autonomous mobile system (AMS), the object's position has to be determined with sufficient accuracy. Especially within a flexible environment its exact 3D pose relative to the manipulator often is not known a priori. The object recognition unit of the presented system robustly accomplishes this 3D pose estimation using a single CCD camera mounted in the gripper exchange system of the mobile robot. The reliability of the results is checked by an independent fault-detection module. A recovery unit handles most of the possible problems autonomously, thus increasing the system's availability.

Often in an industrial environment we meet tasks where technical surfaces have to be machined in relation to natural surfaces or in relation to model textures drawn by stylists and designers. The task is to combine the designed data with the technical constraints coming from the data flow and the manufacturing methods at hand (CAD, CAM). A new method is described that makes use of a laser as a tool. The laser beam is controllable for both its movement in space and its power, that is important for the matter/laser interaction. The combination of the laser with image processing based measurement systems leads to a very flexible system approach for rapid prototyping and small batch manufacturing applications. This new method is described with the emphasis on new areas of application.

In laser welding, problems often arise from the accuracy required by the laser process, particularly where joints have narrow tolerance limits, e.g. with a fillet weld at an overlap joint. In a number of applications seam-tracking sensors can improve this situation. They are able to detect and follow the joint geometry autonomously. In addition to the tolerances, a varying gap between the parts to weld can cause welding flaws. To solve the problems caused by the height of the gap a functionality for adaptive welding can be integrated into the tracking sensor, rendering possible a determined influence on process parameters. Functional dependencies between the height of the gap and the welding parameters are presented in this paper. To further enhance the accuracy of path tracking the dynamic behavior of the system is investigated. With the integration of these dependencies into the tracking sensor, an algorithm for adaptive welding has been obtained, which takes another step towards the raise of profitability of laser installations by a simplified weld seam preparation and an enhanced stability of the welding process.

In view of the competitive situation on the international market, companies are being forced to develop products more rapidly and with less likelihood of errors occurring. In the recent past, product development has been greatly speeded up, above all by computer-aided methods. However, these are not entirely sufficient to achieve a further reduction in product development times. New manufacturing methods such as 'rapid prototyping' (RP) now make it possible to obtain not only computer models but also actual physical patterns in a very early development stage. As RP technologies gained in strength, a demand arose for prototypes in the actual material which it was intended to use for the production article. Using suitable process chains, it is now possible to produce components from various near-series plastics (for example by vacuum casting) and also from metals (e.g. by lost-wax casting or sand casting). At the Augsburg User Center run by Institute for Machine Tools and Industrial Management of Munich Technical University, processes for the rapid production of prototype tools which unite machining methods, RP technologies and molding techniques have therefore been developed.

A two-frequency laser interferometric path measuring system is introduced which has been developed especially for the precision rapid-positioning systems of micro lithographic large-scale instruments and for ultra-precision technology. The modulation frequency of 640 MHz, generated by a highly stabilized He-Ne-laser allows -- in combination with a novel HF signal processing method -- measuring velocities up to 6.4 m/s without 'loosing' even the least increment (10 nm). It is further metrological merit that the least increment is not generated by an additional NF interpolation -- as was usual so far -- but aries from the parallel processing in the HF part.