FactoryNet: A Large-Scale Dataset toward Industrial Time-Series Foundation Models

2026-05-09

AI Summary

This work addresses the lack of a universal pretraining corpus for industrial time-series data, which hinders the development of foundation models capable of cross-device and cross-task generalization. To overcome this limitation, the authors propose the Setpoint-Effort-Feedback-Context (S-E-F-C) unified representation framework, which enables standardized alignment of multimodal industrial data. Building upon this framework, they construct FactoryNet—the first large-scale industrial time-series pretraining dataset—comprising 23,000 real and synthetic task executions. Models trained within this paradigm demonstrate zero-shot cross-device transferability and parameter-efficient anomaly detection, achieving performance on par with high-dimensional baselines across 24 aligned signals. FactoryNet is publicly released and continuously expanded, providing critical infrastructure for foundational model research in industrial AI.

Abstract

We introduce the first universal pretraining corpus for industrial time-series data: FactoryNet. 51M datapoints across 23k end-to-end task executions (13.3k real, 9.8k synthetic) on six embodiments, unified by a shared schema that enables robust zero-shot cross-embodiment transfer and highly parameter-efficient anomaly detection. We introduce a novel schema: Setpoint, Effort, Feedback, Context (S-E-F-C) underlying the whole pipeline that maps any actuated system into a common representational frame. The corpus spans 27 annotated anomaly types alongside healthy baselines and counterfactual pairs across robotic manipulation and machining domains. Cross-embodiment transfer experiments yield positive results: under bias-aware metrics our model demonstrates fair cross-embodiment transfer capabilities on the evaluated source-target pair, while 24 schema-aligned signals achieves competitive anomaly detection performance compared to high-dimensional baselines. We release FactoryNet as a growing, multi-embodiment dataset to drive progress toward industrial foundation models.

Problem

• industrial time-series
• foundation models
• cross-embodiment transfer
• anomaly detection
• pretraining dataset

Innovation

• FactoryNet
• industrial time-series
• foundation models
• cross-embodiment transfer
• S-E-F-C schema