Abstract:While modern 5-axis computer numerical control (CNC) systems offer enhanced design flexibility and reduced production time, the dimensional accuracy of the workpiece is significantly compromised by geometric errors, thermal deformations, cutting forces, tool wear, and fixture-related factors. In-situ sensing, in conjunction with machine learning (ML), has recently been implemented on edge devices to synchronously acquire and agilely analyze high-frequency and multifaceted data for the prediction of workpiece quality. However, limited edge computational resources and lack of interpretability in ML models obscure the understanding of key quality-influencing signals. This research introduces InterpHD , a novel graph-based hyperdimensional computing framework that not only assesses workpiece quality in 5-axis CNC on edge, but also characterizes key signals vital for evaluating the quality from in-situ multichannel data. Specifically, a hierarchical graph structure is designed to represent the relationship between channels (e.g., spindle rotation, three linear axes movements, and the rotary A and C axes), parameters (e.g., torque, current, power, and tool speed), and the workpiece dimensional accuracy. Additionally, memory refinement, separability, and parameter significance are proposed to assess the interpretability of the framework. Experimental results on a hybrid 5-axis LASERTEC 65 DED CNC machine indicate that InterpHD not only achieves a 90.7% F1-Score in characterizing a 25.4 mm counterbore feature deviation but also surpasses other ML models with an F1-Score margin of up to 73.0%. The interpretability of the framework reveals that load and torque have 12 times greater impact than power and velocity feed forward for the characterization of geometrical dimensions. InterpHD offers the potential to facilitate causal discovery and provide insights into the relationships between process parameters and part quality in manufacturing.

PREVIS -- A Combined Machine Learning and Visual Interpolation Approach for Interactive Reverse Engineering in Assembly Quality Control

In situ monitoring for numerical controlled manufacturing of large conceptual prototype based on multi-view stitching fusion

Embedding Visual Cognition in 3D Reconstruction from Multi-View Engineering Drawings

Feature-Based Reverse Engineering Modeler--RE-SOFT

Adaptive acquisition planning for visual inspection in remanufacturing using reinforcement learning

Implementation of 3D Optical Scanning Technology for Automotive Applications

Reconstructing original design: Process planning for reverse engineering

Industrial Vision Inspection Using Digital Twins: Bridging CAD Models and Realistic Scenarios

A rapid generation method of models in machining processes for real-time human–machine interaction with virtual-real fusion

Hierarchical representation and interpretable learning for accelerated quality monitoring in machining process

Intelligent vision based wear forecasting on surfaces of machine tool elements

A Progressive Visual Analytics Tool for Incremental Experimental Evaluation

Robust and High-Performance Machine Vision System for Automatic Quality Inspection in Assembly Processes

VIRL: Volume-Informed Representation Learning towards Few-shot Manufacturability Estimation

Study of CAD Modeling for Reverse Engineering

Animation of virtual mannequins, robot-like simulation or motion captures

A novel adaptive visualization method based on user intention in AR manual assembly

Deep Learning Models for Visual Inspection on Automotive Assembling Line

AutoVis: Enabling Mixed-Immersive Analysis of Automotive User Interface Interaction Studies

Single-View 3D Reconstruction via Differentiable Rendering and Inverse Procedural Modeling

Integrated application model for visual detection of welding quality based on visual neuron under edge-end collaboration