AI-Based Surface Roughness Prediction Model for Automated CAM-Planning Optimization

Tonejca, Lea

doi:10.34726/5450

Record link:

http://hdl.handle.net/20.500.12708/193844
https://doi.org/10.34726/5450

Title:

AI-Based Surface Roughness Prediction Model for Automated CAM-Planning Optimization

Citation:

Tonejca, L. (2023, June 5). AI-Based Surface Roughness Prediction Model for Automated CAM-Planning Optimization [Presentation]. Manufacturing Performance Days 2023 - Workshop “Industry 4.0 for a resource efficient and cleaner production,” Tampere, Finland. https://doi.org/10.34726/5450

reposiTUm DOI:

10.34726/5450

Publication Type:

Presentation - Presentation

Language:

English

Authors:

Tonejca, Lea

Organisational Unit:

E311-01-3 - Forschungsgruppe Steuerungstechnik und integrierte Systeme

Date (published):

5-Jun-2023

Event name:

Manufacturing Performance Days 2023 - Workshop "Industry 4.0 for a resource efficient and cleaner production"

Event date:

5-Jun-2023 - 7-Jun-2023

Event place:

Tampere, Finland

Keywords:

Smart Manufacturing; Industrial AI; Surface Optimization

Abstract:

The surface quality of machined aluminum parts is being optimized by an AI, that feeds on historical structured data of milling and drilling processes to reduce scrap and increase efficiency. The AI output will further be fed back to the CAM software and a live simulation during the machining process itself.

Project title:

Development of Self-Learning NC-Programming System for Data-Driven Surface Prediction and Adaptive Cutting Parameter Optimization: FO999891268 (FFG - Österr. Forschungsförderungs- gesellschaft mbH)

Additional information:

Presentation of the advances of the SurfAIce project, which is funded by the Austrian Research Promotion Agency FFG.

Research Areas:

Mathematical and Algorithmic Foundations: 30%
Modeling and Simulation: 30%
Computational System Design: 40%

Science Branch:

2030 - Maschinenbau: 70%
2050 - Werkstofftechnik: 15%
2020 - Elektrotechnik, Elektronik, Informationstechnik: 15%

License:

In Copyright

Appears in Collections:

Presentation