AI summaryⓘ
The authors introduce a flexible system that can understand and work with different types of time-series data, like sensor readings over time, to help with monitoring and maintenance tasks. Their system uses multiple pretrained encoders to extract useful information and a gate to pick the best encoders for each dataset, combining their insights using a special attention method. This design allows the model to adapt to different tasks like filling missing data, predicting future values, and learning from few examples without retraining everything. Tests show that their method performs well compared to others, and it also works effectively in a real industrial case involving temperature monitoring in hydro-generator rotors.
Digital TwinsPrognostics and Health Management (PHM)Time-series DataSelf-supervised LearningRepresentation EncoderTransformerSelf-attentionImputationLong-term ForecastingFew-shot Learning
Authors
Quang Hung Pham, Ryad Zemouri, Martin Gagnon, Luc Vouligny
Abstract
Engineering Digital Twins and Prognostics and Health Management (PHM) systems rely on robust perception modules to extract actionable information from heterogeneous and non-stationary time-series data. However, most existing approaches remain task-specific, data-hungry, and difficult to integrate into scalable monitoring and decision-making pipelines. Moreover, purely data-driven models often lack robustness and transferability across varying operating conditions. To address these challenges, this paper proposes a modular foundation model for time-series perception based on a collection of pretrained representation encoders. The framework leverages self-supervised learning on heterogeneous datasets to learn transferable and task-agnostic representations, which can be reused across multiple PHM tasks. A gating mechanism is introduced to dynamically select relevant encoders for a given target dataset, enabling conditional computation and adaptive model composition. The selected representations are projected into a shared latent space and aggregated using a Transformer-based self-attention module that explicitly models cross-encoder interactions. The resulting architecture supports multiple downstream tasks, including imputation, long-term forecasting, and few-shot learning, through lightweight task-specific heads, while keeping pretrained encoders frozen during adaptation. Extensive ablation studies demonstrate the complementary roles of self-supervised pretraining, encoder selection, representation alignment, and adaptive aggregation. Experimental results on the ETT benchmark show competitive performance across tasks, while a real-world industrial case study on virtual sensing for hydro-generator rotor temperature highlights the practical relevance of the approach.