Search papers, labs, and topics across Lattice.
To overcome the label scarcity and drift susceptibility of task-specific models in process industries, the authors build IPM-FM, a foundation model architecture tailored for industrial process monitoring and soft sensing. The framework combines self-supervised pretraining on an Informer backbone with multi-criteria consensus feature selection, recursive lag-feature regression, and Monte Carlo dropout to handle asymmetric sampling and safety-critical uncertainty estimation. Evaluated on a seven-year hydrotreater dataset, IPM-FM improves RMSE by 14.6% over from-scratch sequence baselines and 8.3% over classical methods while maintaining 97% empirical coverage at a nominal 95% prediction interval.
Shifting industrial process monitoring from brittle bespoke models to a pre-trained foundation model cuts soft-sensing error by up to 14.6% while guaranteeing calibrated uncertainty across seven years of operational drift.
Industrial process monitoring is fundamental to the safety and economic performance of modern process plants. Current practice remains a one-task-one-model paradigm that is label-inefficient and prone to degradation under operating drift. Foundation models have reshaped language, vision, and generic time-series forecasting, but it has not been adapted to industrial process monitoring. This setting poses domain-specific challenges, including safety-critical decisions and asymmetric sampling between process variables and laboratory measurements. We propose the industrial process monitoring foundation model (IPM-FM). It first learns general-purpose representations from unlabeled industrial process data through self-supervised pretraining, then adapts to specific monitoring tasks using a small amount of task-labeled data, and finally produces calibrated predictions through an uncertainty-aware prediction head. IPM-FM integrates a self-supervised Informer backbone with a multi-criteria consensus feature selector, a recursive lag-feature regression head, and a calibrated Monte Carlo dropout uncertainty module. On a seven-year hydrotreater dataset for diesel flash-point soft sensing, IPM-FM attains an RMSE of 2.99, $R^2$ of 0.50, and 97\% coverage of its 95\% predictive interval, outperforming the strongest classical and from-scratch sequence baselines by 8.3\% and 14.6\% in RMSE respectively, supporting the viability of a unified pretraining--adaptation framework for industrial process monitoring.
