chapter eight

8 Integrating inference and intelligence

 

This chapter covers

  • The importance of a solid data foundation for intelligence
  • Detecting anomalies and faults
  • Forecasting the future with classic machine learning approaches
  • Generative and agentic AI

Visualizing the data that we have collected about the physical system our digital twin represents enabled us to build a descriptive digital twin, allowing us to see the current state of the system and use that information to drive better decisions. In this chapter, we will look at how you can integrate AI and ML into your digital twin to anticipate the system’s future state.

We have looked at how to sense changes in a physical system, communicate, and store them, to build a digital representation of reality through data contextualization. This solid data foundation is absolutely essential before considering how to incorporate AI/ML to build a predictive digital twin—​one that uses historical data to anticipate future states of a physical system.

We begin by looking at classical AI/ML techniques, including supervised and unsupervised learning, and explore how they can be applied to different types of data stored in a digital twin. We then explore generative AI and agentic architectures that extend these capabilities into systems that can observe, reason, plan, and execute actions with minimal human intervention.

8.1 Starting with a strong data foundation

8.2 A practical data pipeline

8.2.1 Cleaning and normalization

8.2.2 Time alignment and resampling

8.2.3 Feature extraction

8.2.4 Storage

8.3 Anomaly and fault detection

8.3.1 Rule-based detection and its limitations

8.3.2 Statistical approaches

8.3.3 Unsupervised learning

8.4 Predictive modeling and machine learning

8.4.1 Training a forecasting model

8.4.2 Applying AutoML

8.4.3 Supervised learning

8.5 Generative AI

8.5.1 Retrieval augmented generation

8.5.2 Agentic AI

8.5.3 Building an agent for the home digital twin

8.5.4 When agents complement or replace control strategies

8.5.5 Multi-agent systems

8.6 Choosing the right approach

8.6.1 Operational viability

8.6.2 Safety and human in the loop