Industrial process model improves monitoring accuracy with fewer labels
IPM-FM: A Foundation Model with Consensus Feature Selection for Industrial Process Monitoring
Machine LearningArtificial Intelligence
Summary
Industrial plants need to keep their processes safe and efficient, but monitoring these processes can be tricky and often requires lots of labeled data. The authors propose a new model called IPM-FM that first learns from large amounts of unlabeled industrial data, then quickly adapts to specific tasks with only a small amount of labeled information. This approach also estimates how confident the model is in its predictions, which is important for safety. On a real-world seven-year dataset, this model predicted diesel quality better than traditional methods.
What this means in practice
- •For process control engineers: Improve monitoring and fault detection in chemical plants by using a pre-trained model that adapts to specific process variables with limited labeled data.
- •For manufacturing quality teams: Generate more reliable quality measurements through soft sensing, reducing dependence on costly and infrequent laboratory testing.
Tested on one dataset.
Authors
Liang Cao, Weide Liu, Yan Qin, Jun Cheng, Weisi Lin, Bhushan Gopaluni
Abstract
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.