AI summaryⓘ
The authors address the challenge of scheduling routes for automated material transport in semiconductor factories, where delays can happen due to congestion at certain equipment. They developed a new method called TN-DCR that uses historical transport data and network information to better predict delivery times and congestion risks before making routing decisions. Their approach combines several data-driven tools to estimate these factors accurately and uses them to choose routes that avoid extreme delays. Testing their system showed it reduced average delivery and waiting times without lowering the number of completed transports. This work helps make factory operations smoother by improving real-time scheduling decisions.
automated material handlingsemiconductor fabricationroute schedulingqueueing theorycongestion predictiontransport networkgraph embeddingordinal classificationempirical Bayesoperations research
Authors
Hao Yin, Meiqi Tu, Anbang Liu, Shaochong Lin, Max Z. J. Shen
Abstract
Automated material handling systems in semiconductor fabs are operated by a material control system (MCS) that must schedule a relay route for every transport command online, before execution. This is a data-driven scheduling problem in which route cost is dominated in the upper tail by queueing at heterogeneous, partially observable relay equipment, so route selection requires estimating both delivery time and congestion risk at the decision moment. This paper proposes a transport-network-aware dynamic congestion representation (TN-DCR). Built on a static directed transport graph induced by historically observed relay segments, TN-DCR combines structural route priors, multi-window network-wide congestion context, route-level bottleneck exposure, and an inductive graph-aware route embedding, all constructed under a prediction-time-safety invariant that admits only information observed strictly before the prediction moment. The representation feeds separate queue- and transfer-time regressors and an ordinal multi-label classifier producing calibrated multi-threshold exceedance scores, with an empirical-Bayes stock-key residual correction reducing systematic queue-time underprediction. The predictions serve as costs in a risk-constrained route-scheduling rule that minimizes predicted delivery time subject to a bound on extreme-congestion probability, embedding the learned predictors within a lightweight operations-research decision model. In a controlled closed-loop evaluation, mean delivery time falls by 16.4\% and internal resource waiting time by 22.6\% while throughput remains essentially unchanged.