Assembly of a Gripper for a Wall-mounting Robot

Main Sector of Application

Manufacturing of robots and machines

Potential Sector/s of Application

Manufacturing of other types of equipment
Solutions for sectors like construction, aviation, food and medical areas

Description of the Experiment

Precizika Metal is a typical SME company that is evolving from being a component supplier to delivering full solutions. This implies an increasing need to deliver fully assembled products to clients. Due to the lack of qualified workforce (or at least being a challenge to find qualified workers) and due to a constant need to be more flexible but also more efficient, methods to automate the current assembly processes that are still performed manually have to be developed.

The Precizika metal (PRZM) use case consists of the assembly of a gearbox. A total of 5 parts have to be assembled using various techniques. Assembly parts enter the cell on a pallet situated on a trolley equipped with a plug and produce (PnP) connector. The PnP connectors developed in the ReconCell project enable trolleys to be quickly and rigidly connected to the workcell. The PnP equipped trolleys are the basic building blocks of the workcell that handle material flow, store application specific robotic grippers and add specialized equipment. In the use case of PRZM a total of four PnP trolleys are used. A trolley is used to store assembly components another stores robotic grippers the last two are used for storing a pneumatic press and a clamp. The assembly procedure consists of the following steps

  1. Robot 1 grasps the roll side support from the assembly component trolley using a parallel gripper. The roll side support is then placed it in the pneumatic press equipped with the appropriate die. Innovations in term of developing a pneumatic press with robotically reconfigurable dies will be considered. Such a system will allow multiple different parts to be assembled using the same press with minimal or no human intervention. A vision system will be used to determine the part type (to prevent operator error) and precise position.
  2. Robot 2 picks up the plain bearing bushing using a parallel gripper and centers it relative to the roll side support.
  3. While the robot 2 still holds the bearing bushing the press imprints the bussing in the roll side support. While the press closes robot 2 follows the movements of the die thus guiding the bushing in place.
  4. Robot 1 grasps the brake shaft and centers it relative to the plain bearing bushing now imprinted in the roll side support. To achieve this robotic reconfiguration of the press might be needed.
  5. While the robot 1 guides the brake shaft in place while the pneumatic press closes.
  6. Robot 1 removes the subassembly and places back on the tray.
  7. Robot 1 then inserts the pitch to roll flange in the clamp. The clamp jaws will be specially designed to precisely position the part while clamping it.
  8. Robot 2 meanwhile changes the end effector to a three finger one. The three finger gripper enables precise positioning of the brake shaft using three flat surfaces 120° apart. The subassembly consisting of the brake shaft the plain bearing busing and the roll side support is then mounted on the pitch to roll flange clamped in the clamp. A milled slot in the brake shaft guarantees proper alignment.
  9. Robot 1 meanwhile changes the end effector to a robotic screwdriver and screws the two m8 screws joining the roll to pitch flange to the subassembly done earlier.
  10. Robot 2 meanwhile changes the gripper and grasps the kendrion base. A three finger gripper is used. It uses three radially placed mounting holes 120° apart as a reference to guarantee proper ordination of the handled part.
  11. Robot 1 joins the kendrion brake to the brake shaft by screwing the three radially placed M8 screws (used earlier as a reference) in the kendriong base. The use of a vision system for detecting the position of mounting holes will be considered.ng
  12. The clamp is opened and robot 2 grasps the finished subassembly and places it on the tray.

Main Objectives of the Experiment

The amount of parts to be produced per year is expected to be between 500-1000 pieces. The production will start in smaller quantities and then be scaled up.

The main objective of the experiment is to show that the ReconCell system can be used to assemble robot grippers and that it is able to deal with

  • Small/medium size batches
  • Variations in gripper design
  • High accuracy of assembly

Expected Results

In this experiment we will develop an automated gripper assembly process for the ReconCell system. The goal is to assemble a robot gripper with minimal human intervention, with at least the same accuracy as a human worker, and with a short enough setup time so that the automated assembly becomes more profitable compared to doing the part assembly manually.

Partial implementation of gripper assembly in ReconCell. The workcell had been reconfigured by manually interchanging modules, like the tool rack trolley, part trolley and press trolley in order to accomodate the assembly of this product.

Final assembly & customization of drive systems and control boxes

Main Sector of Application

Electronics industry
Furniture industry

Potential Sector/s of Application

Part customization

Description of the Experiment

LOGICDATA is an innovation leader and the leading supplier in the field of motor controls, operating elements and actuator elements for electronically adjustable furniture.

Our challenge is that we have high volume products, but with various customization possibilities for our customers. This leads to many different product versions, based on only a few basic versions. The basic versions are produced in high volumes on specific production lines and then manually finalized before packing. The final assembly and customization includes:

  • final housings assembly
  • adding customer specific parts and labels
  • flashing customized software

In this experiment we will test the applicability of the ReconCell system for final assembly & customization of drive systems. The assembly task consists of the following steps:

  1. The preassembled basic version of the drive system comes into the cell packed in a cardboard box. The customization parts, the required tools and grippers are introduced using trolleys. The trolleys connect to the workcell using Plug and Produce (PnP) connectors and provide efficient and precise material flow.
  2. Robot 1 picks up the camera to measure the orientation and distance of adjustable parts of the drive (spindle height, orientation, slimdrive height). Robot 2 picks up the slimdrive, puts it in the position for camera measurements and afterwards inserts it in the assembly fixture.
  3. Robot 1 picks up the screwing tool, while the robot 2 picks up the customization parts. If the slimdrive requires the top plate customization, robot 2 holds it, while robot 1 screws it on top of the drive.
  4. The slimdrive is then rotated for further assembly. If the extension customization is required, robot 1 changes the screw head, robot 2 picks up the part and holds it, while robot 1 screws it onto the drive.
  5. If the tube adapter customization is required, robot 1 changes its tool and picks up the parts. Meanwhile robot 2 adjusts the length of the slimdrive by rotating the stator.
  6. After the assembly of customizations is finished, robot 2 grasps the slimdrive and places it in the appropriate packaging.

Main Objectives of the Experiment

Customization far away from the customers leads to long delivery times, especially for those who are located overseas. The automated robot workcell will give us opportunity to do the final customization close to the costumer in a cost efficient way for all different products.

    Final assembly and customization of all different products
  • Easy and quick change between product groups
  • Throughput that matches our future market demand
  • Easy worldwide setup possible
  • Fast introduction of new variants without knowledge of robot programming

Expected Results

With this experiment we will demonstrate the technical feasibility of robotic assembly and gather data for economic calculations. Based on the results the business case will be validated.

Partial implementation of drive assembly in ReconCell. The workcell had been reconfigured by manually interchanging modules, like the tool rack trolley, part trolley and fixture trolley in order to accommodate the assembly of this product.

Assembly of Automotive Lights

Main Sector of Application

Automotive industry

Potential Sector/s of Application

Manufacturing

Description of the Experiment

The assembly of automotive lights in ELVEZ is currently to a large degree done manually. Typically 8 workers are involved to realize the work in a required cycle time. The production takes place in four shifts (24/7). Manual cell consists of special assembly fixtures, positioning and supporting devices mounted on a frame that forms the assembly devices where the left and right lights are assembled in the same sequence.

In this experiment, we will introduce the assembly procedure using the ReconCell system. The assembly task consists of the following steps:

  1. Light housings (in our experiment we used models x82 and x07)and other assembly components come into the workcell on pallets mounted on trolleys.

    Logo

    X07 light housing

    Logo

    X82 light housing

    The trolleys are connected to the workcell using Plug and Produce (PnP) connectors. These connectors will be crucial not only for material flow but also for quick reconfiguration as they will be used to connect different peripheral equipment e.g. fixtures. The connectors provide power, Ethernet and pneumatic connection as well as mechanical coupling. The position of the assembly components is precisely determined by using 3D vision. so that the robots can easily and reliably grasp the assembly components.

  2. The first robot grasps first the housing part (either x07 or x82). Housing parts will be placed into the special reconfigurable fixtures. The fixtures are based on an unactuated Stewart platform that can be configured for a specific model of the housing. The reconfiguration can be performed by robot thus minimal human intervention in needed. This significantly shortens changeover time making robotic cells feasible for small and medium enterprises that usually manufacture small batches of highly diversified products.
  3. For the assembly of the bulb holder and LWR motor the second robot will use a double revolver gripper. The x07 model requires the assembly of both LWR drive and bulb holder while the x82 model requires the assembly of the LWR drive only. The appropriate assembly components will be picked up from specially designed trays. To ensure that the assembly components are correctly inserted, robot force control will be used for accurate task execution. Force-torque data will be used to define KPIs, which will be evaluated and further transferred to the right simulation and management point.
  4. For automatic screwing we will examine some innovative technical possibilities (e.g. fixed or movable screwdriver with integrated, automatic screw feeding system). However, all variants require robot positioning to the right location, and fine-tuning the position through the robot force control.
  5. The X07 light housing model requires the insertion of a sheet metal part that acts as a heat shield and prevents overheating of heat sensitive plastic parts. The first robot will perform this task. Using a magnetic gripper, the heat shield part will be picked up from a palette and slid in place.
  6. After finishing the assembly, the first robot grasps a camera and inspects the assembly quality control points. When the inspection is done, the second robot puts the light into the package for further transportation.

By making use of the business intelligence system integrated with the ReconCell system, we will be able to show real benefits of the proposed reconfigurable assembly system from the order to delivery.

Main Objectives of the Experiment

When companies introduce a new production process, they do feasibility studies to estimate its profitability. Often companies (95%) take into account just the quantity of parts when the production process is in full run, but they forget that after the end of regular production, spare parts must still be produced. This means that companies must take into account that first they will be making a large number of parts, but after the end of regular production they will still need to make a smaller number of parts.

For every new automotive light (start of production), ELVEZ has to order two new assembly devices (left side, right side). After the end of production, assembly devices must not be disassembled because they are needed to manufacture spare parts. Assembly devices must therefore be stored in a company for the next 10 years. This means that company needs a lot of free space to store these assembly devices. Production of spare parts is a low quantity aproduction and is typically done only once per year. Assembly devices are not universal but are different from product to product.

The purpose of the experiment is to show that we can use a reconfigurable robot workcell to assemble different products, thereby alleviating the need to store many different assembly devices. For example, with the same injection moulding machine, we will be able to make different headlight housings with typical structural elements for three different lights. One light is at the end of production, the second is in full run, while the production of the third light is expected to start during the lifetime of ReconCell project.

Expected Results

With this experiment we will demonstrate the applicability of the reconfigurable robot workcell on a concrete example from the production of ELVEZ, i.e. assembly of automotive lights.

Implementation of the X07 and X82 light housing assembly in ReconCell. The workcell includes passive fixture modules called hexapods, that allow active reconfiguration by the robot. After assembling the first housing, the workcell autonomusly reconfigures itself in order to assemble a different housing.