Precision Assembly Technologies and Systems

The assembly of optical components goes along with highest requirements regarding assembly precision. Laser products have become an integral part of many industrial, medical, and consumer applications and their relevance will increase significantly in the years to come. Still economic challenges remain. Assembly costs are driven by the demanding requirements regarding alignment and adhesive bonding. Especially challenging in precision bonding are the interdependencies between alignment and bonding. Multiple components need to be aligned within smallest spatial and angular tolerances in submicron order of magnitude. A major challenge in adhesive bonding is the fact that the bonding process is irreversible. Accordingly, the first bonding attempt needs to be successful. Today’s UV-curing adhesives inherit shrinkage effects during curing which are crucial for the submicron tolerances of e.g. FACs or beam combiners what makes the bonding of these components very delicate assembly tasks. However, the shrinkage of UV-curing adhesives is not only varying between different loads due to fluctuations in raw materials, it is also changing along the storage period. An answer to this specific challenge can be the characterization of the adhesive on a daily basis. The characterization before application of the adhesive is necessary for precision optics assembly in order to reach highest output yields, minimal tolerances and ideal beam-shaping results. The work presented in this paper aims for a significantly reduced impact of shrinkage effects during curing of highly durable UV-curing epoxy adhesives resulting in increased precision. Key approach is the highly precise volumetric dispensing of the adhesive as well as the characterization of the shrinkage level. These two key factors allow most reproducible adhesive bonding in automated assembly cells. These proceedings are essential for standardized automated assembly solutions which will prospectively play a major role in laser technology.

A new assembly method for silicon micro-parts with non-planar steel parts is proposed: low-temperature soldering. Existing techniques for micro-mechanical assemblies are analyzed and compared to the proposed method, the method is explained and validated on an existing product using functional tests. Performances of the reference method (adhesive bonding) are not yet fully reached, but the results are close and partially fulfil the specifications of the current production.

The superposition of ultrasonic frequency vibrations to conventional machining techniques is known and practiced since the 1950s under the name of ultrasonic machining. Using ultrasound, many good properties appear including reduced thrust force, improved surface finish, reduced residual stress in machined material, etc…In this paper we present a new assembly technique based on the same principles: ultrasonic press–fitting. Feasibility and energy reduction are demonstrated through experiments under industrial production conditions.

The work presented in this paper aims for a significant reduction of process development and production ramp up times for the automated assembly of micro-optical modules and systems. The approach proposed bridges the gap between the product development phase and the realization of automation control through integration of established software tools such as optics simulation and CAD modeling as well as through introduction of novel software tools and methods to efficiently describe active alignment strategies. The focus of the paper is put on the formalized specification of product configurations as a basis for the engineering of automated assembly processes. The concepts are applied to industrial use-cases. The paper concludes with an overview of the application of the concepts in an engineering tool-chain as well as an outlook on the next development steps.

The feeding of components smaller than 1mm² is a high challenge in an automated manufacturing line. The surface forces affecting these micro components are getting more important and cause problems when handled. A new method is described using the surface forces for separating, sorting and arranging micro components. The surface shape of fluids and the gravity are used for moving the floating components to defined positions. The components can be arranged in a magazine or can be sorted by a channel system.

The paper focuses on the study of strategies and tools to handle miniaturized components in the electronic industry. In particular, the paper presents an innovative device and method to manipulate microcomponents by vacuum. The device includes an original releasing system, that does not require any external actuation, to assist their release. Indeed, at the microscale, adhesion forces predominate over the gravitational force due to the small masses of the microcomponents, often leading to the failure of the release phase if a release strategy is not implemented. The device, able to eliminate the adhesion problem, is compared with a traditional vacuum microgripper in terms of grasping and releasing error and percentage. The results of preliminary experimental tests are discussed, demonstrating that the innovative microgripper represents an interesting solution for handling electronic components as well as different microparts.

This paper describes the dynamics of capillary self-alignment of components with initial shift offsets from matching receptor sites. The analysis of the full uniaxial self-alignment dynamics of foil-based mesoscopic dies from pre-alignment to final settling evidenced three distinct, sequential regimes impacting the process performance. The dependence of accuracy, alignment time and repeatability of capillary self-alignment on control parameters such as size, weight, surface energy and initial offset of assembling dies was investigated. Finally, we studied the influence of the dynamic coupling between the degenerate oscillation modes of the system on the alignment performance by means of pre-defined biaxial offsets.

Image stitching is a method that forms a bigger image out of two or more smaller images with a certain overlap in their field-of-view. Usage of this technique is present in photography, medical imaging and other fields as well. The potential for using image stitching for measurement purposes in industrial applications has not been investigated as thoroughly, but it is receiving more attention. This paper examines the possibility for creating a machine vision-based measurement system that employs commercially available image-stitching platforms for measuring the length of long medical screws. Three methodologies were tested for performing medical screw measurements and this paper summarizes the findings of the image-stitching approach.

Highly individualized and customized products with dynamic lifecycles increase the need for flexible and reconfigurable assembly systems. Industrial robots are a key technology for future production systems especially for large scale components. The trade off between increasing work piece dimensions and constant or even increasing tolerance requirements, that are in some cases comparable to micro assembly systems, has to be solved by flexible and precise manufacturing and fixtureless assembly processes.

Precision mechanism is widely used for various industry applications. Quality inspection for precision mechanism is essential for manufacturers to assure the product leaving factory with expected quality. In this paper, we propose a novel automated fault detection method, named Tilear, based on a Deep Belief Network (DBN) auto-encoder. DBN is a probabilistic generative model, composed by stacked Restricted Boltzmann Machines. With its RBM-layer-wise training methods, DBN can perform fast inference and extract high level feature of the inputs. By unfolding the stacked RBMs symmetrically, a DBN auto-encoder is constructed to reconstruct the inputs as closely as possible. Based on the DBN auto-encoder, Tilear is structured in two parts: training and decision-making. During training, Tilear is trained with the signals only from good samples, which enables the trained DBN auto-encoder only know how to reconstruct signals of good samples. In the decision-making part, comparing the recorded signal from test sample and the Tilear reconstructed signal, allows to measure how well a recording from a test sample matches the DBN auto-encoder model learned from good samples. A reliable decision could be made. We perform experiments on two different precision mechanisms: precision electromotors and greasing control units. The feasibility of Tilear was demonstrated first. Additionally, performance of Tilear on the acquired electromotor dataset was compared with the state-of-the-art machine learning based fault detection technique, support vector machine (SVM). First result indicates that Tilear excels the SVM in terms of the Area Under the Curve (AUC) obtained from the Receiver Operating Characteristics (ROC) curve plot: 0.960 achieved by Tilear, while 0.941 by SVM.

Colloidal lithography based on the assembly of microspheres is a powerful tool for the creation of a large variety of two dimensional micro or nanostructures patterned. However very few studies examine the control, qualification and quantification of the ordering of the particles once deposited on the substrate. We have developed two unique methods working at microscopic and macroscopic scales, respectively called Microfixe® and Macrofixe®, for the analysis of grain morphology in the case of hexagonal closed packed (HCP) monolayers of spherical microparticles. The processing of the images taken at microscopic scale uses Delaunay triangulation and histograms of lengths and orientations of Delaunay triangles sides. At the macroscopic scale, six camera images are required of the sample illuminated under six different incidence angles. Image treatment consists in the comparison of the six images and eventually subdivision of these images to sharpen the analysis. At the end, the two softwares constitute artificial images of particle deposit, giving at microscopic and macroscopic scales significant information about grain size, grain morphology, orientation distributions, defaults (voids, stacking)…With these two new control methods, colloidal lithography is emerging as an industrial process.

Due to constant globalisation new trends on the market are coming up. One of the trends is the customisation of products for the customer and shorter product life cycles. To overcome the trends industries identified as key element self-reconfigurable production systems. A change to a running system means loss of time, money and manpower. A reconfigurable production system can automatically adapt to changes in terms of changing a machine or a product. The methodology behind is adapted from the office world and is called plug and produce. However, a production system has different requirements which need to be met. Due to a lack of homogeneity of industrial controllers in terms of communication and reconfigurability, as well as the interaction with the end user, the multi-agent technology was identified as a superior communicator. We present a new multilayered multi-agent architecture where the necessary agent types are introduced to fulfil the requirements for plug and produce. One scenario is shown where the architecture is employed to enable plug and produce capabilities and allow the system to adapt itself.

The production environment in Europe is characterized by a number of different challenges. Production volumes are hardly predictable due to the turbulent market [1]. Product life cycles become increasingly shorter [2]. Customers demand more individual and customized products, which leads to increasingly diversified product pallets [3]. The occurrence of changes on product and process, even for running productions, increases continuously [4]. At the same time the field of tension between assurance of quality, reduction of lead times and reduction of cost remains present [5].

In micro production high lot sizes, which are required for cost-effectiveness, are often not achieved, e.g. due to a high number of product variants or limited demand. Therefore new approaches for production like changeable manufacturing systems are demanded. In this paper a concept for a modular workpiece carrier system including a platform for sensors and actors is presented. The modular design allows the easy adaption of the manufacturing system to new products by the workpiece carrier system in terms of hardware and even the integration of additional process functionality.

Manufacturing is getting more competitive with time due to continuously increasing global competition. Late market introduction decreases the economic lifecycle of products and reduces return on investments. Reconfigurable Manufacturing Systems (RMS) reduce the time to market because the process of equipment configuration is less time consuming than engineering it from scratch. This paper presents a scientific framework, to be applied as an engineering design tool, that is capable of improving the relation between product design and the reconfiguration process of RMS. Not only does it support a cross-domain adjustment and information exchange between product developers and manufacturing engineers, it also adds risk analysis for conscious risk taking in a cyclic development process. The method was applied on an industrial case; concurrent design and manufacturing of an environmentally friendly circuit board for wireless sensors. The method may be considered successful. It will lead to better system architecture of product and production systems at a more competitive cost. Feedback on the development process comes available in the early development stage when the product design is not rooted yet and two-way optimisations are still possible.

Generative manufacturing technologies are gaining more and more of importance as key enabling technologies in future manufacturing, especially when a flexible scalable manufacturing of small medium series of customized parts is required. The paper describes a new approach for design and manufacturing of complex three dimensional components building on a combination of additive manufacturing and e-printing technologies, where the micro component is made up of stacks of functionalized layers of polymer films. Special attention will be paid to the “3-d” modeling approach, requested to support the applicaton developer through provision of design rules for this integrated manufacturing concept . Both, the application concept as well as the related equipment and manufacturing integration currently are currently developed further in the project SMARTLAM, funded by the European Commission.

Remote Center Compliance (RCC) devices are passive devices used in automated assembly. For round peg-in-hole insertions, initial misalignments of the peg can cause heavy radial loadings on the corresponding hole in arbitrary direction. RCC devices with rotational symmetry property are therefore desirable. This paper discusses how such devices may be designed and shows that circular periodic structures satisfy this property. The elastic part of the RCC device is formulated as a compliant mechanism and a systematic design methodology is proposed based on structural optimization. A smooth optimal design is achieved with distributed compliance and numerical simulation is conducted to illustrate the feasibility. The proposed method is expected to be used to design a novel device for assembling fragile plastic parts, which is a challenge to 3C industry.