Delay-guided attention improves quality prediction in manufacturing
Delay-Guided Attention for Quality Prediction in Multistage Manufacturing
Machine Learning
Summary
Predicting product quality in factories with multiple production steps is hard because delays between steps can change over time. The authors introduce a method called DAMF that learns how long these delays are and uses that information to better guess the final quality. DAMF uses special attention techniques to match delayed data from earlier steps with current stages and improves prediction accuracy without needing fixed delay info from the factory. They tested their method on real and simulated datasets and showed it works better than other approaches.
What this means in practice
- •For industrial process engineers: Improve prediction of final product quality by explicitly modeling varying delays between manufacturing stages for better production control.
- •For manufacturing data scientists: Develop models that use learned temporal delay distributions to align data from different production phases and enhance forecasting accuracy.
Authors
Jiarui Zhang, Zhaoyang Guan, Zhixuan Xia, Naichen Shi
Abstract
Quality prediction in multistage manufacturing is challenging when inter-stage lags vary over time, as measurements taken at the same timestamp may reflect different process states. Existing approaches often preserve production order but fuse timestamp-aligned features or model delayed dependencies implicitly, without identifying the upstream history relevant to each downstream observation. We propose Delay-Guided Attention for Multi-stage Forecasting (DAMF), which explicitly estimates inter-stage lag distributions and uses them to guide cross-stage information transfer for final quality prediction. For each adjacent stage pair, a lag predictor compares the current downstream representation with candidate lagged upstream representations. An occurrence-magnitude factorization separately estimates the probability of a positive lag and its conditional magnitude, yielding a sample- and stage-pair-specific distribution over discrete lag states without requiring plant-specified delays at inference. DelayAlignment incorporates this soft distribution as a temporal prior into differentiable cross-stage attention, while gated residual fusion regulates information propagation across production stages. For interpretation, segment-wise Viterbi decoding converts pointwise lag distributions into temporally coherent lag paths, whereas forecasting retains the soft distributions for alignment. On the real-world LiquidSugar dataset, DAMF improves final quality prediction over the compared baselines. Evaluations on two controlled-delay benchmarks show more accurate and stable lag recovery, and ablation studies support the contributions of lag prediction, delay-guided attention, and structured temporal decoding.