Integrating 4 use-cases


Use-case: Collaborative TV assembly

In Arcelik Çerkezköy Electronics Factory, TVs are assembled. In this usecase we focus specifically on electronic card (PCB) assembly, where two operators take the PCBs from boxes near them and screw the cards in place. In this operation, the operators need to turn a lot and also do lots of repetitive movements which is not ergonomically ideal. To solve this issue, in CoLLaboratE we have designed a collaborative cell containing one collaborative robot (KUKA iwaa) and an operator. First, the worker teaches the robot kinesthetically how to take the PCB from the box and places the card in its place on the TV. Then the robot can execute this task autonomously by detecting the TV using machine vision and collaborate with the the operator who screws the cards in place.

In Arcelik we value our workers and want them to work as safe as possible, this new collaborative cell is a good improvement in ergonomy. We are happy to introduce our operators new ways to interact with robots, it is a chance to broaden our perspective on collaborative robots and the possibilities come with them.

This use-case demonstrates novel functionalities functionalities developed by 5 CoLLaboratE partners, such as teaching by demonstration and execution with movement primitives (AUTH), visual demonstration and extraction of behavioural trees (CERTH), gesture recognition of the worker for robot commanding (ARMINES), operation planning (LMS) and collaboration of with mobile robots (ASTI).

A worker teaching the robot kinesthetically, on how to assemble the two boards on the TV chassis.

Use-case: Collaborative aircraft structure riveting

Ever since the dawn of the all-metal aircraft we are all used to see crossing our skies, riveting has been a largely manual operation. In particular, percussive riveting requires two human operators working in close collaboration: one operator performing the riveting, the second operator producing a resistive force for allowing the plastic deformation of the rivet. This means high vibration, very uncomfortable body postures, loud noise… In a nutshell, a real challenge for ergonomics. And now imagine an aircraft structure requiring one million rivets, or even more. You wouldn’t want to put yourself in the place of the technicians doing the riveting, right?

Well, the good news is that you don’t have to! We now have robots to ease the job, integrating functionalities from 5 partners. The robotic system developed in CoLLaboratE project can work collaboratively with a human technician. This way, percussive riveting becomes a semiautomated process. While the human operator inserts the rivet into the hole, the robotic system identifies this action (CERTH) and places the bucking bar on the rivet head (IDIAP). The human then performs the riveting. Next, the robotic system captures an image of the deformed rivet so that the human can assess the quality of the operation. If the quality is satisfactory, the human instructs the robot to proceed to the next rivet; otherwise, the instruction is to remain on the current rivet for a new riveting session.

Not to mention that the human can communicate with the robotic system by voice (ASTI), by gestures (ARMINES) and/or by means a graphical interface (CERTH). Moreover, the robotic system is extremely flexible. It can change position and height, thus being usable for a variety of structures, different in size and shape. Isn’t this convenient? It looks like the future of industrial assembly has just started.

A mobile manipulator ready to collaborate with a human in riveting.

Use-case: Collaborative assembly of car parts

The vision of the project was to develop a robotic system that will assist or even replace the worker when he is not present at the work cell. Note that in a modern production, humans normally assist more than one work cells and it is less likely that robots would tightly cooperate with humans all the time. Therefore, the worker and the robot must be both capable of executing same operations, where the work sharing is assigned dynamically.

The process of insertion is demanding due to the high flexibility and elasticity of the sliding rings. Previous attempts of automation failed for many reasons, but mainly because it’s not possible to assure the required success rate of the insertion into the fixtures. For this purpose multiple approaches and technologies were developed and used to facilitate this functionality.

The main technological innovations and improvements that we have developed within the project and applied to the above mentioned use case are:

  • human intention recognition based on RGBD camera and recurrent neural networs (RNN),
  • dynamic task sharing between robot and the operator,
  • assembly of object with low tolerances supported by multi modal exception strategy learning and ergodic control,
  • automatic cell adaptation to fit the individual human ergonomics based on passive reconfigurable hardware.

All these innovative technologies were integrated into the cell in such a way that it does not require special operator skills for operation.

Close up of the workcell with the human teaching the robot (on the right) how to assemble the flexible copper components in the mould.

Use-case: Collaborative inspection and pre-assembly of a car windshield

In the CRF use-case, the human has to inspect and assemble components on a heavy, cumbersome and fragile car windshield. The use-case doesn’t aim at the effective modification of an existing line, but it is focused on the realization of a new workcell. The two CoLLaboratE technologies that allow the human to work together with a 1000kg robot, are the sensorized handles that can distinguish the human touch (UNIGE) and the variable admittance control (AUTH) that allows easy and precise manipulation of the 100kg payload.

This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 820767.

The website reflects only the view of the author(s) and the Commission is not responsible for any use that may be made of the information it contains.

Project Coordinator
Prof. Zoe Doulgeri
Automation & Robotics Lab
Aristotle University of Thessaloniki
Department of Electrical & Computer Engineering
Thessaloniki 54124, Greece
Collaborate Project CoLLaboratE Project
© 2018-2022 All rights reserved.