Lehrstuhl für Mechatronik in Maschinenbau und Fahrzeugtechnik (MEC)

KORINS: Komplementäres Robotergestütztes Inspektionssystem für Mehrwegflaschen

Problem Formulation

In the beverage industry, after the introduction of uniform glass bottles as well as so-called modular cases, a tightening of individual marketing to the neglect of uniform glass bottles and cases has been clearly pronounced and implemented. When returning the empty deposit bottles, there is a large mixing of different types of bottles and individual bottles in the cases. Some manufacturers therefore carry out a manual re-inspection to ensure that more and more bottles remain in the deposit pool. KORINS is developing an automatic sorting system that can be integrated into existing systems and, based on AI methods, inspects and sorts the bottles and cases. The research focus is on the development and implementation of algorithms for cooperative planning of robotic tasks for optimal sorting performance and the development of a digital twin for online system monitoring and maintenance.

 

Solution Approach

The complementary sorting system mainly consists of two inspection stations, the case inspection followed by a 3D inspection of the glass bottles. During the case inspection, a visual system consisting of cameras and ultrasonic sensors is used to examine and analyze the bottle-case affiliation in order to identify bottles that are highly unlikely to correspond to the current case (foreign bottles). In the second inspection station, immediately before sorting, bottles with unclear affiliation are again individually inspected. Based on this, a set of objects (bottles to be sorted) and slots (free or incorrectly occupied positions in the cases) is defined and passed on to the scheduling algorithm. The scheduler plans the tasks for the robots and passes them to the MPC algorithm for trajectory planning. The generated robot trajectories are then passed to the local robot controller via ROS. A Digital Twin and a cloud service are used to create a digital representation of all system components, with the goal of developing a system for online monitoring of all components and their dependencies.

 

Project Goals

  • Mathematical modeling of the robot task allocation and scheduling involving multiple objects representing a variety of bottles and cases
  • Model based planning and control of cooperative collision-free trajectories for multiple robots sharing a common workspace.
  • Development and implementation of the image processing software and the machine learning algorithms for bottle inspection.
  • Design and implementation of a Digital Twin for the sorting system and Integration of the cloud services.
  • Development of an experimental setup consisting of a lightweight robots, stereo cameras, ultrasonic sensors and a digital twin.

 

KORNIS Architecture

Keywords

  • Cooperative Robotics
  • Digital Twins
  • Scheduling

 

Funding

Projektträger: AiF-Projekt GmbH

 

Time span

Sep 2020 – Mar 2023

 

Project Partners

Ximaj IT-Solutions GmbH

 

Contact

Prof. Dr.-Ing. Naim Bajcinca
Gottlieb-Daimler-Str. 42
67663, Kaiserslautern
+49 (0)631/205-3230
naim.bajcinca(at)mv.uni-kl.de

Zum Seitenanfang