用小模型融合三类特征，实现工业设备异常提前预测。
Triplet Feature Fusion for Equipment Anomaly Prediction : An Open-Source Methodology Using Small Foundation Models
- 将统计、时序和文本特征融合成1116维向量，输入轻量分类器
- 30天预测精度达99.2%，误报率降低83%至0.1%
- 全流程可在本地CPU运行，适合无网络环境部署
在工业设施管理中，提前预测设备异常至关重要。现有方法或依赖人工设定阈值（泛化性差），或使用大型神经网络（难于现场部署）。本文提出一种开源方法，将小规模基础模型融合为统一的1,116维三重特征融合管道：(1) 基于90天传感器历史的统计特征（$R^{28}$），(2) 通过LoRA微调的IBM Granite TinyTimeMixer（TTM，133K参数）生成的时序嵌入（$R^{64}$），(3) 用multilingual-e5-large从日文设备台账中提取的多语言文本嵌入（$R^{1024}$）。三者拼接后输入轻量级LightGBM分类器（<3 MB），预测30、60、90天后的异常。所有组件均采用宽松开源协议（Apache 2.0/MIT）。推理仅需2毫秒内完成，可在边缘硬件上独立运行。在包含64台空调机组、共67,045个样本的数据集上，30天预测的精确率（Precision）达0.992，F1为0.958，ROC-AUC为0.998。关键突破是误报率从基线0.6%降至0.1%，下降83%，归因于文本嵌入对设备类型的条件建模。聚类分析显示，嵌入能将时序特征与特定故障模式对齐，说明紧凑的多语言表示可提升判别力而无需显式类别编码。
原文摘要 · Abstract (English)
Predicting equipment anomalies before they escalate into failures is a critical challenge in industrial facility management. Existing approaches rely either on hand-crafted threshold rules, which lack generalizability, or on large neural models that are impractical for on-site, air-gapped deployments. We present an industrial methodology that resolves this tension by combining open-source small foundation models into a unified 1,116-dimensional Triplet Feature Fusion pipeline. This pipeline integrates: (1) statistical features (x in $R^{28}$) derived from 90-day sensor histories, (2) time-series embeddings (y in $R^{64}$) from a LoRA-adapted IBM Granite TinyTimeMixer (TTM, 133K parameters), and (3) multilingual text embeddings (z in $R^{1024}$) extracted from Japanese equipment master records via multilingual-e5-large. The concatenated triplet h = [x; y; z] is processed by a LightGBM classifier (< 3 MB) trained to predict anomalies at 30-, 60-, and 90-day horizons. All components use permissive open-source licenses (Apache 2.0 / MIT). The inference-time pipeline runs entirely on CPU in under 2 ms, enabling edge deployment on co-located hardware without cloud dependency. On a dataset of 64 HVAC units comprising 67,045 samples, the triplet model achieves Precision = 0.992, F1 = 0.958, and ROC-AUC = 0.998 at the 30-day horizon. Crucially, it reduces the False Positive Rate from 0.6 percent (baseline) to 0.1 percent - an 83 percent reduction attributable to equipment-type conditioning via text embedding z. Cluster analysis reveals that the embeddings align time-series signatures with distinct fault archetypes, explaining how compact multilingual representations improve discrimination without explicit categorical encoding.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。