Reproduce “Improving Industrial Injection Molding Processes with Explainable AI for Quality Classification” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Georg Rottenwalter, Marcel Tilly, Victor Owolabi

Research from Rosenheim Technical University of Applied Sciences demonstrated how Explainable AI (XAI) can identify critical process parameters in industrial injection molding, leading to improved quality classification with a reduced set of sensor inputs. Using XAI-guided feature selection, models achieved higher accuracy and faster inference, making AI-driven quality control more accessible for manufacturing systems with limited sensor capabilities.

Key insights

  • XAI methods, particularly SHAP and LIME, consistently identified 'Injection Pressure,' 'Actual Clamping Force,' and 'Screw Position' as highly influential features for quality classification.
  • Reducing input features based on XAI analysis can enhance model generalization, with the 9-feature model outperforming the full 19-feature model in accuracy.
  • XAI-driven feature reduction simplifies the integration of AI quality control into environments with limited sensor data and can streamline the generation of synthetic training data.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.