Students: Siebe van den Berg, Anthony Luzac, Pascal de Vilder and Marieke Bakker (2015)
Alumat Zeeman is a family company, founded in 1928. The company designs, manufactures and innovates greenhouse parts, vent opening systems and screen installations for the international horticulture sector. [1]
One of the company’s fields of expertise is wire processing. A lot of different products are being made using a wire processing machine, the Wafios; think of coil-like products.
A group of four students from The Hague University of Applied Sciences were asked to help the company with a packaging problem. In short, their task was to figure out a way to get a wire processed product from the wire processing machine, using a robot (Universal Robot, UR5), to a packaging station.
The original situation
Alumat Zeeman is working with an automated machine (figure 1) for its truss clamps (figure 2). The wire processing machine, the Wafios (figure 3), cooperates with this automated machine. The automated machine receives a wire product by a pneumatic Wafios arm with beaks and places the product onto a product carrier (figure 1).
| Figure 1, automated machine placing products onto a product carrier (white strip) |
| Figure 2, truss clamp (left) and netting bracket (right) |
Alumat Zeeman doesn’t use the automated machine for its netting brackets (figure 2). The netting brackets fall into a box and are sent to a sheltered workshop, where personnel sorts out the wire products and makes sure they are placed onto a product carrier (figure 7).
The problem
When receiving a truss clamp with the pneumatic Wafios arm, a force is applied to cut off the wire product. This is the moment when the product sometimes slips out of the gripper of the pneumatic Wafios arm. As the product moves out a bit, the next step, to place the product onto the automated machine, has a high chance of not being able to be fulfilled. The system receives an error and stops the production.
| Figure 3, the Wafios |
Apart from optimizing the automated machine’s situation, Alumat Zeeman is also interested in automating the situation of the netting bracket since these are placed on product carriers by the personnel at the sheltered workshop which costs extra money for transportation and time delay.
The solution
A solution to the problem described above is to use a robot. The E.O.A.T. (End Of Arm Tool) should be flexible in the direction the Wafios cuts off the wire products to be able to deal with this force. Therefore, leaf springs are used, being able to bend for at least one centimeter dealing with the force. In combination with the so called fingertips (red cap, figure 4 and 5) of the E.O.A.T. a flexible tool arises. Two sets of fingertips for the E.O.A.T. are designed in order to grab both the truss clamps (figure 4) and the netting brackets (figure 5). These sets of fingertips can be changed within one minute to minimize the time left without any production (changeover time). Figure 4 shows the truss clamp being grabbed by the E.O.A.T. during a test being done at Alumat Zeeman. A slot is applied and in combination with strips of another material being glued to the aluminum leaf springs, unwanted vibrations of the tool can be controlled. Due to the use of leaf springs and a thin layer of rubber on the fingertips, the truss clamp stays firmly fixed between the fingertips. Figure 5 shows that a thin layer of rubber also makes sure that the netting bracket stays in place when being held by its E.O.A.T..
| Figure 4, truss clamp grabbed at the Wafios; slot (left side) |
| Figure 5, netting bracket grabbed during in-lab test |
In order to know if the product is actually grabbed, a sensor is applied. This sensor detects whether or not a product is grabbed by means of conductivity. When a product is grabbed it closes the electrical circuit. The product is then detected.
The products should then be stacked onto a product carrier. These product carriers were kept as how they were received from Alumat Zeeman because this method turned out to be the best after doing research (figure 6 and 7). The product carrier holders, holding the product carriers, are shown in figure 8 as red holders. These holders can easily be removed from the profile they are attached to.
| Figure 6, the truss clamp carrier |
| Figure 7, the netting bracket carrier |
The robot’s script was written using the coordinates from its environment, dealing with the signals it gets from the Wafios, using the sensor’s information for product counting and using a timer function to measure the cycle time. The script was optimized to achieve the shortest cycle time. The script itself can be edited relatively easy for another size of a product family; for now this would be either the truss clamp’s product family or the netting bracket’s product family.
Decisions made during the project
Using a robot for this project comes along with some main decisions to be made. Apart from the information given in the description of the solution, some other decisions have influenced the design process as well.
First of all, when the product carriers (figure 8) are filled with products, they need to be carried away. This part of the process is out of the scope of the project making it possible to focus on the other parts of the process.
Second of all it was decided in agreement with Alumat Zeeman that by the end of the project a working prototype for stacking truss clamps onto a product carrier could be shown at Alumat Zeeman’s factory. In addition it was also decided in agreement with Alumat Zeeman that by the end of the project a working proof of concept could be shown to Alumat Zeeman at The Hague University of Applied Sciences to stack the netting brackets onto a product carrier.
In order to be able to show the prototype and the proof of concept, several in-lab tests needed to be done. The environment near the Wafios was to be built as close to the real environment as possible (figure 8) to make the in-lab tests work. During the in-lab tests the signals of the Wafios were simulated using a microcontroller. The microcontroller was programmed to send a signal to resemble the Wafios, enabling the in-lab tests without the need of the real machine.
Last but not least, in agreement with Alumat Zeeman, a cost analysis was made to compare the original situation, using the automated machine, and the new situation, using the robot.
| Figure 8, product carriers and the environment |
Concluding numbers
Applying a robot instead of the automated machine raises the production by 16,67% truss clamps per hour. When focusing on costs and gains the robot pays itself back approximately 9,5 months faster than the automated machine. The robot is therefore recommended to replace the automated machine.