KaiwuDB TSFMs (Time-series Foundation Models)
With the core framework of 'Data Governance - Analysis - Decision-making', drives the popularization of temporal intelligence and helps IoT enterprises rapidly achieve digital and intelligent transformation.
Core Features
Automation
Automation is the key to make large-scale industry models more accessible and easier to use. In scenarios where there are numerous new energy power stations, a wide variety of energy sources (including solar, wind, and ocean energy), and a broad geographical spread, traditional methods require developing multiple specialized models for different situations, which can be extremely costly. KaiwuDB TSFMs can automatically select and customize models for specific scenarios without manual intervention, significantly save the cost of model development and maintenance.
Generalization
Generalization unlock the invisible walls that stood between different scenarios and datasets. Through pre-training on time-series data across multiple domains, the model gains strong zero-shot/few-shot cross-task capabilities, enabling rapid adaptation to a wide range of smart + applications, such as Digital Energy, Intelligent O&M, Smart Emergency Response, Smart Cities and more. The model can flexibly respond to various scenarios, without the limitation of traditional models that require customization for each specific case.
Robustness
Robustness ensures the model's stable output in complex environments. In practical application, time-series data often disturbed by noise and streaming transmission. By targeted optimizations, the model can effectively handle above problems, while address data missing and outliers, thereby guarantee the reliability of analytical results.
Explainability
Explainability makes AI decision-making more trustworthy. The model's analytical results align with physical laws and can clearly present basis behind its decisions. For instance, when equipment failure warnings appear, it can pinpoint specific issues such as abnormal temperature or pressure fluctuations, rather than just providing general conclusions. This is crucial in industrial O&M and equipment early warning scenarios.
Efficiency
Efficiency ensures diverse deployment requirements. The lightweight model can support large-scale cloud data processing, and also can apply to resource-constrained edge devices, enabling rapid reasoning analysis to guarantee real-time decision-making.
Data Governance - Analysis - Decision-making Framework
Data Governance
KaiwuDB TSFMs centers around the time dimension to build a powerful data governance framework. When dealing with multi-source time-series data (e.g., temperature, pressure, rotational speed) in industrial scenarios, and multi-model data from sensors(e.g., pressure, flow rate, temperature) in urban pipeline networks, KaiwuDB TSFMs perform well in data integration, missing value imputation, and uncertainty quantification. This effectively enhances data quality and provides efficient data accessibility.
Data Analysis
Data analysis is the core ability of time-series intelligence. KaiwuDB TSFMs supports a full range of time-series analysis tasks across various scenarios, including prediction, classification, anomaly detection, clustering, and Q&A. For instance, it can detect abnormal rotational speeds in real time based on equipment operational data; while predict the probability and exact time of geological disasters by comparing historical and current rainfall data. These analytical capabilities apply to discrete and continuous data, meeting complex business scenarios.
Data-based Decision-making
KaiwuDB TSFMs take a step forward from 'data insights' to 'actionable guidance', which empower Agentic AI to make decisions in personalized, multi-objective, and uncertain contexts. For instance, it can pinpoint the exact timing for shutdown maintenance or parts replacements based on equipment lifecycle predictions, making decisions more targeted and actionable.
