Evidence record · UDINUS · SDGs
- Unit
- Rektorat UDINUS
- Period
- 2025
- Goals
- Record
- ccda9499-4dae-4227-a42d-e22bbd1d4cd4
- Published
- 2026-10-30
Universitas Dian Nuswantoro conducted an activity on machine learning-driven optimization for predicting surface roughness in vertically oriented 3D printed components. The activity addresses the challenge of obtaining accurate surface roughness measurements in Additive Manufacturing, where traditional point-based measurements may fail to capture the full range of surface characteristics. The activity involved Islahudin, Nur; Nugroho, Dony Satriyo; Wijaya, Dewa Kusuma; Amalia; Suprijono, Herwin; Ginta, Turnad Lenggo; Azka, Muizuddin; and Rahadian, Helmy as participants.
The activity investigated the use of machine learning to improve roughness prediction by integrating image-based analysis with real-time adaptive control. A full factorial experimental design was used to examine the effects of infill density, print speed, nozzle temperature, and layer height on the surface roughness of vertically oriented printed parts. The evidence describes a dataset of 81 experiments, including roughness image data, which was used to test and validate an XGBoost model.
The proposed model produced stronger results than the traditional regression approach. The machine learning model achieved an R2 of 97.06% and a Mean Squared Error of 0.1383, while the conventional approach achieved an R2 of 95.72% and a Mean Squared Error of 0.224. Within the scope of the activity, these results demonstrate the potential of machine-learning-driven optimization to improve precision and consistency in Additive Manufacturing. The source relates this potential to Additive Manufacturing applications in aerospace, healthcare, and automotive industries.
In relation to Sustainable Development Goal 9, the activity provides evidence of research and innovation focused on improving an Additive Manufacturing process through image-based analysis, adaptive control, experimental design, and XGBoost validation. The activity therefore contributes to the documented institutional framing of innovation and infrastructure research at Universitas Dian Nuswantoro during 2025.