AUTOMATION AND ROBOTICS:APPLICATIONS AND CASE STUDIES

ASSEMBLY IN INDUSTRY: APPLICATIONS AND CASE STUDIES

Assembly is fundamental in almost all industrial activities. Assembly applications in several important industries are illustrated in the following case studies.

Automotive Assembly

More than half of the industrial robots throughout the world are used in the automotive industry. These robots are utilized in a number of different manufacturing sectors. The main areas of application are in car-body construction, mainly in spot welding and painting. Assembly lines are also being more and more robotized in the automotive industry. This is essential in the case of complex joining processes requiring a high standard of quality, which can be guaranteed only by robots.

But in order to promote automation in automotive assembly effectively, two conditions must be met first:

1. Assembly-oriented automotive design

2. Preassembly of as many subunits as possible

A few application areas of automatic assembly in the automotive industry (see Figures 35, 36, 37) are shown in the following examples.

Assembly of Large Steering Components

This non-synchronous pallet-based, power and free system manufactures a large steering component for the automotive industry (see Figure 38). Cycle time is approximately 20 seconds, three shifts per day. Unique pallet design helped this customer automate the assembly process while minimizing material handling of the final product. An important issue for the manufacturing process was the testing of each unit throughout the assembly sequence. Tests performed are leak, torque, function, and final performance. The automation system also featured manual stations for placing specific parts and for rework and evaluation.

Automatic Removal of Gearboxes from Crates

The combination of innovative sensory analysis with modern industrial robots is opening up more new fields of application for automation technology. Automatic unloading of workpieces from crates is a problem that automation technology researchers have been addressing for a long time. However, only recently have usable systems for the rough production environment been developed. The re- quirements for these unloading systems are:

• Localization accuracy better than 5 mm

• Cycle time maximum 24 sec (in accordance with subsequent manufacturing processes)

• No predefined lighting conditions for the recognition sensors

• Lowest possible tooling-up requirements for new types of workpiece

• If possible, no additional pieces of equipment for the recognition sensors in the work cell

In the case outlined here, the processed and partly sorted cast parts are automatically unloaded from the crates. In order to avoid the cast parts damaging each other, they are stacked in the crates on cushioned layers (see Figure 39).

After the filled crate has been delivered to the work cell, the robot first measures the four upper edges with the help of a 2D triangulation sensor that is attached to the robot. This measurement is used to calculate the exact location and height of the crate.

The robot sensor is guided over the workpieces in the crate to determine the position of each of the workpieces. The workpiece surface that has been recognized is used to determine the three positions and orientations of the workpiece. The approach track with which the robot can grab and pick up the cast part is prepared from the workpiece position. The whole approach track is calculated in advance and reviewed for possible collisions. If the danger of a collision exists, alternative tracks are calculated and the next cast part is grabbed if necessary. In this way the system maintains a high level of operational security and has sound protection from damage.

Electronic Assembly

Assembly of an Overload Protector

This pallet-based assembly system produces a safety overload protection device and was designed with multiple zones (see Figure 40). The transport system is a flex-link conveyor with a dual-tooled pallet for cycle time consideration. The cycle time for the system is 1.2 sec. Tray handling and gantry robots were utilized for this application due to the high value of the component. The line features unique coil winding, wire-stripping technologies, and sophisticated DC welding for over 20 different final customer parts for the electronic assembly market.

Assembly of Measuring Instruments

The production of measuring instruments is often characterized by small batches, short delivery times, and short development times. Automated assembly of the measuring instrument cases enables prod- ucts to be manufactured at lower costs. The cases are made of several frame parts. Each frame part is separated by a metal cord that provides screening against high-frequency radiation.

The different kinds of cases have various dimensions (six heights, three widths, and three depths). Altogether there are about 1,000 product variants with customer-specific fastening positions for the measuring instrument inserts. Therefore, the assembly line is divided into order-independent preas- sembly and order-specific final assembly (see Figure 41).

In the preassembly cell are automatic stations for pressing in different fastening nuts and screwing in several threaded bolts. An industrial robot handles the frame parts. After removing the frames from the supply pallets, the robot brings the frame into a mechanical centering device for fine

positioning and the elimination of tolerances in the pallets. In this way it is possible to achieve very exact positioning of the frame in the robot gripper.

During the pressing operation, the robot positions and fixes the frame in the press station (press force of about 3000 N). A compliance system is integrated in the gripper to eliminate tolerances in dimensions of the frame. The robot also positions and fixes the frames in the screwing station. There is a system to control the rotation angle, screwing torque, and screwing depth to reproduce a screwing depth of 0.1 mm.

After finishing preassembly, the robot places the frames on a conveyer system. The conveyer links the pre- and final-assembly cells. The final-assembly cell work steps are (see Figure 42):

• Fitting the metal cord for high-frequency screening into the frames

• Order-specific pressing of fastening elements

• Screwing together all frames to produce the finished case

• Lettering the finished case

The difficulty in assembling the metal cord is that these parts have no rigidity and can only transmit tractive forces. Two metal cords with different diameters (2.0 mm and 3.0 mm) have to be fitted into four fundamentally different running slots. To achieve the required stability of the cord in the rear-frame slot it is necessary to put in adhesive points. A cord-cutting system is integrated into the robot tool to achieve the right length of the cord (depending on the dimensions of the frames), and an adhesive dispensing system is integrated into the tool for placing the adhesive spots into the slots.

After the metal cord has been fitted into the different frames (front inside, front outside, rear, and side ledge), the fasteners for the inserts are pressed in order-specific positions into the side ledges. For this operation the robot takes the ledge with the required length out of a magazine and brings it into a press station. A guide rail defines the exact position. The fasteners are blown automatically from the feeder through a feed pipe to the press position.

The subsequent screwing of all frame components is also executed by the robot. The robot moves to the screwing position. The screws are blown automatically through the feed pipe from the feeder to the screwing tool. The rotation angle and screwing torque are monitored during the screwing operation to achieve a perfect result.

Depending on the construction of the case, it is important to have accessibility from four directions during the complete assembly process. Therefore, a clamping device that can turn the case in all

required positions is essential. The clamping device is made up of standard components that can clamp more than 25 cases with different dimensions. The robot arm has an automatic tool-changing system for changing tools quickly; each tool can be picked up within a few seconds.

Experience with the assembly cells outlined above has shown that even assembly tasks with extensive assembly steps producing less than 10,000 units per year can be automated in an efficient way.

Assembly of Luminaire Wiring

For the optimal economic assembly of luminaire wiring, a simultaneous development of new products and process techniques was necessary. A new method developed for the direct assembly of single leads has proved to be a great improvement over the traditional preassembly of wires and wiring sets. Single leads are no longer preassembled and placed in the end assembly. Instead, an endless lead is taken directly from the supply with the help of a newly developed fully automated system and built into the luminaire case. This greatly simplifies the logistics and material flow inside the company and reduces costs enormously.

The IDC technique has proved to be the optimal connection system for the wiring of luminaires. Time-consuming processes like wire stripping and plugging can now be replaced by a process com- bining pressing and cutting in the IDC connector. The introduction of this new connection technique for luminaire wiring requires modification of all component designs used in the luminaires.

Fully automatic wiring of luminaires is made possible by the integration of previously presented components into the whole system. Luminaire cases are supplied by a feeding system. The luminaire case type is selected by the vision system identifying the bar code on the luminaire case. Next, luminaire components are assembled in the luminaire case and then directed to the wiring station (Figure 43).

Before the robot starts to lay and contact the leads, a camera, integrated in the wiring tools, is positioned above the required position of the component designated for wiring. Each component is identified and its precise position is controlled by vision marks in the shape of three cylindrical cavities in the die casting of the IDC connector. Any deviation of the actual position from the target position is calculated and transmitted as compensation value to the robot’s evaluation program.

Quality assurance for the processes, especially for pressing and contacting of the conductor in the IDC connector, is directly controlled during processing. When the wire is inserted into the con- nector, a significant and specific force gradient is shown within narrow tolerances. This measured gradient can be easily compared and evaluated immediately after pressing with the help of a reference

graph. In the case of nonconformity, the luminaire is isolated, taken out of the material flow system, and directed to a manual work station for rework.

Depending on the geometry of the luminaires, it can be assumed that the application of this assembly system can reduce the share of wage costs for each luminaire by up to 65%.

Assembly of Fiberoptic Connectors

While communications and information technology are rapidly evolving, transmission media require- ments are increasing more and more. The enormous increase in data rates, the increase in data transmission speeds, and the high level of reliability of EMI mean that copper-bounded transmission technology has reached its technical and economical limits because of its susceptibility to interference and limited transmission rates. Fiberoptic technology, by contrast, has already shown its potential and advantages in telecommunication and computer network (LAN) applications. Because manufac- tured units are increasing, automation of the fiberoptic connector assembly is becoming a more and more exciting field.

Fiberoptic cables are currently assembled almost completely manually. Semiautomatic devices are available on the market for the premanufacturing process. After this process is completed, the ceramic ferrule of the connector is filled with glue, then the glass fiber is introduced into the ferrule hole. This assembly sequence is executed only manually at present. However, because no measuring sys- tems are available for this purpose, manual assembly cannot ensure that the critical joining force will not be exceeded. If this happens, the fiber will break off. The fiber will be damaged and the connector will become waste.

Figure 44 shows the construction of an ST connector and the simplex cable that is used.

An automated fiberoptic connector assembly, especially in the inserting process, is generally obstructed by following factors: glass fiber sensitivity to breaks, nonrigid cable, complex cable con- struction, inserting dimensions in the �J.m range, small backlash between fiber and ferrule (0–3 �J.m), varying glue viscosity, and maximal inserting forces in the mN range.

Figure 45 shows a prototype assembly cell where the above-mentioned subsystems are integrated.

Microassembly

Increasing miniaturization of mechanical components and rising demands in the precision mechanics and microtechnology industries are creating high demands on assembly automation. Today automated solutions only exist for mass series production systems (the watchmaking industry, the electronics industry). Automation of assembly tasks for small and medium-sized series is still rare. Flexible automated microassembly systems are required there due to the large number of different types and variants.

The assembly of precision mechanics and technical microsystem components often requires join- ing work to be conducted in the micrometer range. Intelligent robot grabbing systems have been developed in order to achieve the join accuracy required. Motion-assisted systems are particularly suited for flexible offsetting of tolerances in short cycles.

Figure 46 shows an example of a oscillation-assisted joining robot tool for precision and microas- sembly of small planetary gears. Assisted by the adjustable oscillation in the grabber and the regulated tolerance offset, each of the gearwheels and the gearwheel housing can be process assembled very securely by means of the integrated fine positioning.

The special features of the joining tool are the joining accuracy of up to 2 �J.m that can be achieved, the process-integrated offsetting of orientation and position deviations (up to ±1 mm), and the high level of adjustment as well as flexibility essential for performing the different joining tasks. The robot tool is used especially for phaseless bolt-in-hole joining tasks with accuracy requirements in the micrometer range and for the assembly of complex toothed wheel pairs.

Food Industry

The number of convenience products sold annually in recent years has increased markedly. Particu- larly in everyday life and in cafeterias the trend is toward meals that can be prepared more quickly and conveniently. According to the experts at Food Consulting, the share of convenience products in

meat products will increase from 15% at present to above 30% in the next two years. The food industry is changing into a preparation industry and its vertical range of manufacture will increase dramatically. This will necessitate the introduction of automated manufacturing processes and thus enable the use of new automation and robot systems.

In addition to increasing productivity, automated systems allow improved levels of hygiene and thus extending the sell-by date of food products. They also allow ever-increasing consumer demands for more transparency regarding product origins to be met.

Figure 47 shows a robot in use in the food-packaging sector. At the end of a production line, the unsorted sausages that arrive on the conveyor belt are recognized by an image-processing system and automatically sorted according to the final packaging unit with the help of the robot.

Pharmaceutical and Biotechnological Industry

In the pharmaceutical industry, large sums of money are used for preclinical and clinical research. The development of a drug costs millions of dollars. Hundreds of tests have to be conducted. One of the innovative tools in the field of drug discovery is microchip technology. Microchips require less reagent volume, make analytical processes run faster because of their smaller size, and allow more sensitive detection methods to be implemented. In this way, they reduce costs, save time, and improve quality. Microchips come in two main categories: chips based on microfluidics, with com- ponents such as pumps, mixers, microinjectors, and microarray chips, which have numerous locations of samples, such as DNA samples, on their surface. Most of the interest now seems to be focused on this second category.

Figure 48 shows an arrayer for low-volume spots. With a capillary-based tip printing method, volumes in the lower pl-range can be produced. These very small drop volumes give the opportunity to use more sensitive methods of detection and produce high-density arrays. The automated arrayer is based on a gantry robot. Its three axes are driven by electronic step drives that have a positional accuracy of ±1 �J.m and a speed up to 300 mm / sec. The work area is 300 X 300 mm and the total dimension is 500 X 500 mm. In order to have a controlled environment, the system is operated in a clean room.

So far, commercial available arraying systems are useful only for R&D applications and research labs. They have not achieved wide commercial availability.