Projects

The fact that our camera systems are also suitable for the harshest environments is proven by our recently installed 3D system for the inspection of raw cast gear shafts.

senwork develops AI-based technology that enables 100% quality assurance of LabDisks or Lab-on-a-Chip designs in the production cycle. High precision manufactured, transparent plastic cartridges can now be recorded in the production cycle with this inspection station and reliably evaluated.

Camera systems with telecentric lenses and the corresponding working distance are often too long to be integrated into a machine without space problems. No problem with the Cognex In-Sight based VisionUnit from senswork.

Reading best-before dates on yoghurt pots is a challenge for classic image processing systems, as the writing on a curved background is often distorted and varies in size and contrast. With the help of deep learning, letters and numbers can be read reliably in different light and background conditions.

Classic image processing quickly reaches its limits, especially in the area of ​​surface inspection. With the help of the Neuralzye® software from senswork, an AI inspection of transparent strip material has now been implemented, with which the detection of bubbles and inclusions can be detected.

In an automatic circuit board assembly process, a wide variety of circuit board connectors are automatically pressed with the circuit board. In order to ensure that the circuit board connector can be plugged in in advance, our camera system checks the pin geometries of the plug in advance and uses the pin positions to determine whether it can be plugged into the circuit board image.

For the high-volume production of touch displays, current trends are clearly pointing in the direction of hybrid bonding. This combines the advantages of solid and liquid bonding to permanently connect glass and display. To ensure that the alignment of the individual components is precise, senswork offers high-precision camera systems for determining the correct position and rotation.

Applications in the semiconductor industry require both high-precision, leading-edge technologies and reliable performance characteristics in cleanroom and control test environments. One of the largest manufacturers of semiconductor device handling systems relies on a specially developed VisionUnit from senswork to ensure chip positioning for very small chip sizes with extremely small dimensions.

This project is the first time senswork has implemented the new test equipment concept for highly integrated manual test stations. A high-resolution camera system measures pin positions of an MTD connector assembly, records connector coding and verifies housing dimensions within a tolerance range of 0.1 mm. A color dot spray system marks well-tested components.

Quality assurance is the top priority in the automotive industry. The error tolerances are low. With the ZScan 3D multi-sensor system from senswork, large assemblies made of aluminum or other sheet metal, e.g. oil pans, can be measured three-dimensionally and compared against CAD data.

Volume determination on the production line plays a major role in the food industry. It must be ensured that exactly as much product is packed as is printed on the packaging. With our 3D multi-sensor system ZScan, we can scan cut-to-size or block-format goods such as meat, fish or cheese without shadows and thus determine the exact quantity.

Fakra connectors are used in automotive electronics for the transmission of signal data such as GSM, GPS or camera data. In order to meet the high quality requirements even in the fully automated production of high volumes, a leading supplier of these products relies on senswork‘s inspection technologies.

Membrane tubes are the core component of an innovative filter technology for wastewater treatment plants. To ensure that these tubes are flawless and that the cleaning performance of the plants runs optimally, a line scan camera system from senswork inspects the surfaces of the membrane tubes for a wide variety of production defects.

Plug connections are essential in electrical engineering. Whether in the smartphone, in the car or in production facilities, they ensure that data or electricity flow from A to B. For this to work flawlessly, the connector pins must fit perfectly on the circuit board. With the virtual connector gauge from senswork, the smallest deviations can be detected before damage occurs.

Today, die-pressed carbons and graphites are used in large quantities to manufacture a wide variety of subcomponents for the automotive industry, such as fuel and water pumps. To ensure that the products leave the production line free of defects, senswork has developed a fully automated system for the quality assurance of such components together with a mechanical engineering company from Rosenheim.

In the production of clay bricks, it is critical to ensure that the bricks are dimensionally accurate. This prevents bricks from leaving the factory beyond the manufacturing tolerances and, under certain circumstances, costly return transport from the construction site. For the reliable detection of defects, senswork has developed the BrickChecker, which can be seamlessly integrated into an existing production line.

The gap between contact surfaces plays a major role in connector assemblies. The distance between the individual contact surfaces is measured to an accuracy of 0.01 mm. For this purpose, the test parts are brought into the VisionUnit. The plug is clamped via a clamping mechanism and light diffusers are simultaneously delivered to the assembly, which display the gap between the contact surfaces in high contrast. A high-resolution camera in conjunction with telecentric optics then measures the gap between the individual contacts.

The testing device, which is 100% manufactured by senswork, enables the measurement of Fakra plug assemblies with an accuracy range of ± 0.02 mm. Both the Fakra interface side and the PCB side with press fits are tested with this device. An automatic focusing device ensures that the pins relate to the reference feature with constant image quality. The system is calibrated by a checkerboard calibration with high-precision calibration targets.

When individual components of a multitool cutting tool are fed in, the senswork camera system detects the position and rotation in a flexfeeder feed. Pick candidates are automatically recognized on the Flexfeeder surface on the basis of differently illuminated individual images and the position data is transmitted to the robot. The standard system has been expanded to include additional dark field lighting for reliable detection of the flipped position of all components.

For the automatic joining process of display assemblies with a so-called metal carrier, a total of 6 camera units from senswork support the robot in aligning the components with one another. This is made possible by a unique prism module, which guarantees a precise view in two directions for each camera. This makes it possible to compare the reference point of a display with that of the metal carrier in a camera image.

Our semi-automatic test system measures different Fakra connector assemblies for correct coding, wobble circuit as well as pluggability of the PCB side. To check the pluggability, we use our virtual plug gauge, which recognizes the pluggability in a virtual board image based on the given pin positions.

A dual camera system with mirror deflection enables the reading of part numbers on so-called turbocharger guide vanes by means of OCR. For this purpose, two Cognex In-Sight 8000 smart camera systems were integrated into a senswork VisionUnit with mirror deflection.

In order to check the quality of screw caps on filled soup glasses after filling, senswork now offers a 360 ° camera inspection system, which allows a view of the surface of the cap in a picture through a specially developed mirror deflection. With up to 6 glasses per second, the quality of the lid surface is ensured and checked for damage and freedom from scratches.

The turnkey test system consisting of a camera for the PCB side and a camera for the interface side tests a wide variety of Fakra plug assemblies. This is made possible by an adjustable viewing direction of the camera on the interface side and a well thought-out interchangeable set of partial exposures. Good parts are marked with a needle head that is integrated directly into the part receptacles.

There are many dimensional data on saw blades that ultimately determine the cutting quality and service life. For the inspection of diamond saw blades, senswork, together with the company Plattner, has equipped a measuring station with a double-head laser scanner system, which measures and records all tooth geometries, blade width, cutting width, angularity, and diameter during the run.

For a French automotive supplier, senswork developed and implemented optical control modules for the quality inspection of automotive reversing sensors. Different sensor types are processed in one machine, then packed and stacked in different trays. The modular machine consists of four assembled modules connected by a common control system. A total of seven camera systems are used for optical inspection and defect detection.

In this application, plastic injection molded parts fed by a vibratory conveyor are transferred to a rotating glass plate, where they are inspected by two camera systems. The first of the two systems checks all parts for type accuracy, while system 2 subsequently detects the smallest burrs on the diameter.

With 18 images in different positions and focal planes on the component, our camera system checks a connector assembly on the front and back. With a resolution of 8 µm / pixel and the specially developed stitching tool, the images are joined together and combined into one overall image.

With USB connectors there are different contact surfaces on different Z-planes. In order to ensure a sharp view of the contact quality even at high resolution, this VisionUnit records several images on different focal planes in the shortest possible time. The characteristics are calculated across screens in the test program.

Our camera system inspects automotive labels in a robot cell from a distance of more than 1 m. With a powerful spot lighting, the label can also be illuminated intensively and still homogeneously from this distance. The test program checks texts via OCV, the quality of data matrix codes using AIM-DPM and the quality of printed symbols 100 percent.

With another VisionUnit from senswork with mirror deflection and manual focus adjustment, hundreds of connector assemblies are fully tested every day in an automatic testing machine from Adaptronic. The additional lighting ensures the best lighting conditions for the high-contrast representation of the test features.

In a labeling machine, the product labels are applied with up to 20 applications per second. An incorrectly inserted type of label or incorrectly applied labels can mean recalling entire batches or laborious re-labeling as long as the error is noticed in good time. Barcode scanning systems provide a remedy here: Our barcode technology based on Cognex Dataman reads up to 20 barcodes per second effortlessly and with absolute reliability.

An automotive supplier manufactures connector strips for the specialty "autonomous driving". These strips must be extremely precise and should be tested using an inline process. The ZFokus from senswork enables pin heights to be determined with an accuracy of ± 0.01 millimeters.

As part of a fully automated production, one and the same system should test various connector geometries such as connector strips, Fakra connectors or other HF connectors and their labels. Features such as pin positions, wobble circles and coding features are to be checked. As part of the label inspection, textual content is automatically read and compared using OCV, and symbols and graphics are checked for completeness and print quality / legibility.

Manufacturers of connector components are confronted with ever increasing demands on the precision of the plug connections and their compactness. The ZFokus from senswork makes full use of its advantages in the coplanarity inspection of SMD assemblies and in the pin height check.

Using the example of the soldering point inspection, ZFokus allows the determination of the soldering volume, the detection of gaps in the soldering point and the evaluation of the shape of a soldering point. ZFokus shows its strengths wherever 2D systems today deliver excessive pseudo rejects when evaluating soldered joints due to fluctuating contrast variances or failing to reliably detect defect features.

In order to obtain maximum flexibility with regard to the most varied connector types for manual connector inspection at a manual test station, we use our powerful VisionUnits from senswork with variable focus adjustment. The focus variation is made possible via a 30 mm servo axis, which can be controlled either via a PLC or directly via the camera program.

For a follow-up process in the production of airbag assemblies, it is important to fully automatically recognize bore features on a wide variety of component types and to pass the positions on to a robot. Through an intelligent combination of dark field lighting and senswork circle detection algorithms, this process is 100% supported in production.

Enormous quantities of battery cases are produced every day in battery production for e-cars. Our camera system checks the battery case holder for possible damage, breaks and cases that have not been removed in order to be able to secure the subsequent fully automatic refitting.

Before individual components of an automotive assembly are assembled, they are charged with grease in an automated production line. Here it is important to ensure that the amount of lubricant is applied sufficiently and precisely.

With the manual mounting of assembly groups in particular, it often happens that individual components are missing or the precision of the assembly does not meet the specifications. An end-of-line test system from senswork checks the correct assembly and alignment of individual components.

Our cameras precisely detect the position and angular position of the elements so that assembly elements can be precisely joined in switching elements of electric stoves. The measured data is transferred to a production line with several robots.

In high-frequency technology, the quality of the installed coaxial cables, especially in the connection area of the cables, plays a major role in the subsequent signal quality. The measuring system developed by senswork records the inner conductor tip three-dimensionally, and is thus able to precisely measure the spatial deflection of the tip as well as the protrusion from the outer conductor shield.

The image documentation system records products in resolutions of up to 25 megapixels and allows products to be identified with an optionally available barcode scanner. An automatic exposure time control ensures that the image quality is always constant. A scripting plug-in also enables user-specific actions to be carried out after the image has been taken.

Numerous errors can occur in the manufacture of soaps, and while the human eye can recognize that immediately, the way they look can vary widely. A multi-camera system consisting of incident light, transmitted light and 3D inspection reliably detects all defect features.

In a fully automated handling system with a 6-axis robot, a 3D laser scanner captures components in flight. The robot presents three components per second to the scanner. This records the component geometry in 3D and measures a recess on the component at several positions.

A robot performs mechanical post-processing on cast parts fully automatically. To do this, the robot must know the rotational position of the component to within a few tenths of a degree before the machining process. Only through a specially developed dark field lighting with a correspondingly large diameter could the relevant features for detection be reliably mapped. A Cognex In-Sight smart camera system then determines the rotational position using an edge-based pattern search and forwards this to the robot controller.

Using a 4-fold laser scanner setup, metal rods are checked for correct surface and geometric tolerances. With a 3D calibration, the laser scanners deliver 3D data sets in the same calibrated coordinate system. This method can also be applied to continuous materials. The measuring system is used to measure metal rods and to check the surface for flaws. It is moved over the rod at a speed of up to 100 mm / second.

After the painting process of synthetic parts, they are subjected to a fully automatic surface inspection by a shape-from-shading system before further processing. This method allows the smallest surface defects such as paint pimples, scratches, inclusions, lint or other raised or recessed defects to be reliably detected. For this, our system uses eight lighting directions with eight partial images, which are offset by Shape from Shading to form a height map of the surface.

In this system, Euro pallets are fed via a roller conveyor belt to a robot system, which automatically removes the sack layers and places them on a longitudinal conveyor belt. Using a time of flight camera from senswork, the positions of the plastic bags are recognized fully automatically and transferred to the robot coordinate system. The detection accuracy of the system is ± 10 mm, and thus allows precise pick-up by the suction gripper mounted on the robot.

The cables are 10 km long and tow large ships across the ocean to discover new oil deposits. A defective cable can cause millions in damage if it causes a ship to abandon its business. senswork image processing takes care of this. Even during production, a line camera is swiveled radially around the seam, creating a development of the weld seam. In the camera image generated in this way, the smallest defects in the weld seam are visualized and recognized in good time.

The supplied parts are removed from the supply line by a Scara robot and placed in a moving tray. Our camera technology controls the rotational position of the parts in the feeder, the position of the tray and the correct placement of the products in the tray.

The transmitted light assembly designed for this application measures just 2 mm in material thickness on the illuminated surface. This allows the lighting to dip behind the welding spoons to be measured and to capture and measure their position in an extremely stable manner in transmitted light. We use an In-Sight smart camera system from Cognex as the camera system.

100 percent inline testing of seven-segment modules in production. After the injection molding process, the test parts are checked for the smallest inclusions and inhomogeneities in a double nest in transmitted and reflected light. The senswork inspection tool compensates for nest and shape variances fully automatically and thus ensures an extremely reliable inspection.

This double-head inspection system checks cutting contacts from two directions at the same time by means of cleverly deflected beam paths. The beam path is deflected in such a way that both the interlacing of the contact elements with one another and the gap size can be detected at the same time. A 3D portal system positions this VisionUnit above the contact elements. This VisionUnit saves a complex and wear-prone mechanical test.

A 360 ° panorama endoscope is used to examine the inside of an aluminum body for processing marks and surface defects such as cavities and pores. Our system checks cylindrical internal surfaces in several height sections using deep learning tools for irregularities in the surface.

Chocolate is difficult to describe completely with a rule-based algorithm, since no two chocolates are identical. The individual chocolates have a high natural range of variation.

The individual elements of a first aid kit must be identified through a transparent packaging. Since the packaging reflects light, the individual components of the packaging are difficult to recognize. ViDi Check inspects the complete first aid kit and determines whether all trained features are present.

Varying the font and character size on a non-homogeneous background is very difficult for conventional OCR. With a robust training set, ViDi Read can accurately read letters in a variety of lighting and background conditions.

Varying opacity and bubbles in the liquid create a very inconsistent background in the bottle. The imprinted date and batch codes are difficult to recognize for ordinary image processing. ViDi Read can read letters accurately under a wide range of light and background conditions.

Different liquids create inconsistent backgrounds under the transparent bottle. Plastic seams and difficult reflection angles pose another challenge. ViDi Read can accurately read letters under diverse light and background conditions.

An inhomogeneous topology, different shades and the changing shape of the tortilla make the detection of cracks or wrinkles extremely difficult. ViDi Detect can distinguish between tolerable deviations and defects by training on a reference set of approximately 100 different tortilla samples.

A curciut board for LEDs is very complex and can contain a large number of errors. This makes it difficult to describe all the required features of a good part with rule-based image processing. ViDi Detect uses an example-based approach to model the full range of required circuit board features.

The transparency of glass vials makes it difficult to distinguish breakage from dirt along the front or back. Rule-based image processing reaches its limits here. ViDi Detect can accurately distinguish the variation of natural parts from anomalies by performing a reference set of good parts under varying light conditions and amounts of dirt.

Badly positioned, incorrect or bent components on a populated circuit board are difficult to detect with conventional image processing. Using Deep learning, classification and defect detection is much more reliable, even with a low even with a small amount of training data.

Before spark plugs are packed, they are inspected for completeness per layer. The space situation only permits the use of wide-angle optics, which leads to a strong distortion of the perspective. In this case, classic pattern searches can already reach their recognition limits. Using Deep Learning, however, the the inspection works very reliably.

Defects in air filters must be detected on a complex structured surface. Fluctuations in the light reflection make the use of rule-based image processing impossible. ViDi Detect accurately detects errors in the structured area - despite a high natural variance of the air filters.

Stamped metal at different angles of reflection presents a challenge for rule-based image processing: No two features look exactly alike,especially on curved surfaces. ViDi Read can accurately read letters under a wide range of lighting conditions.