The system allows the recognition of electromechanical components positioned along a DIN rail or inside an electrical panel using an algorithm based on AI for hole identification and matching with the information in the database for the calculation of the their position

Patent title METHOD FOR THE RECOGNITION, IDENTIFICATION AND AUTOMATIC LOCATION OF WIRED COMPONENTS INSTALLED IN A SWITCHBOARD
Thematic area Industry, Digital and Security
Ownership ALMA MATER STUDIORUM - UNIVERSITA' DI BOLOGNA
Inventors Davide Chiaravalli, Alessio Caporali, Kevin Galassi, Gianluca Palli
Protection Italy
Licensing status Available for development, option, licensing and other enhancement agreements
Keywords Electrical Panel, Localization, Vision System
Filed on 12 July 2023

Classical vision systems do not allow the recognition of a generic class of components such as electromechanical components. In addition, the union of several individual terminal blocks can create a potentially infinite combination of components that the vision system must be able to recognize. This invention allows to abstract the figures of the components and recognize the latter through a generalization of all the components having a similar structure with considerable advantages in terms of strength

The vision system proposed is based on the identification of the holes on the upper face combined with the information of the component to be identified allowing the correct recognition of potentially never seen electrical components and in different industrial scenarios (colours, lights, etc.). TRL6

  • Quality control or wiring of components
  •  Applicable in different industrial scenarios
  • The invention allows to recognize with precision (<1mm) the exact position of the components placed inside the frame or along a DIN rail. This information is not known initially due to the manual assembly of the components.
  • Robustness: The system is capable of operating under varying conditions with the same precision, for example different lighting conditions.
Page published on: 31 July 2023