• Conference
  • Engineering and Numerical Tools

Augmented Perception: Empowering Flexible Manufacturing Systems through the Digital Twin – A Novel Approach.

Authors : Yassine FEDDOUL (LINEACT), Nicolas RAGOT (LINEACT), Fabrice DUVAL (LINEACT), Vincent HAVARD (LINEACT), David BAUDRY (LINEACT)

Conférence : Communications avec actes dans un congrès international - 05/08/2024 - IEEE Conference on Industrial Electronics and Applications

In the context of the Industry of the Future, manufacturing environments must be flexible and reconfigurable to continuously adapt to customers’ personalized demands and
changes in the manufacturing processes. This adaptation involves the reconfiguration of the layout and the integration of new systems into the environment: production lines, manufacturing machines, robotic arms, mobile robots, etc. This integration must occur seamlessly within the manufacturing processes to maintain consistent production rates. Currently, there are two options: simulation, which might not fully encompass all parameters of a real scenario; and intervention in the physical system, which requires a production pause and may not guarantee the anticipated benefits of these changes.
To address this issue, we propose to evaluate a new concept
entitled augmented perception, which aims to bridge the real and the virtual to handle new production situations. This evaluation is done through a use case entirely conducted in a simulated environment involving the Tiago++ robot: Tiago++ moves from point A to B and encounters a reconfiguration of the environment by adding an obstacle on its trajectory. Three experiments have been conducted: the first deals with Tiago++ navigation without any reconfiguration of the environment; the second considers the reconfiguration of the environment and the integration of the obstacle in the initial path of the robot without spreading this information to Tiago++’s navigation stack; the last refers to the augmented perception, which involves the digital model of the environment in which this virtual obstacle is integrated. Then this data is spread as contextual information to the robot before navigation occurs.