ScaleOut Model Development Tool™

The ScaleOut Model Development Tool is a GUI development environment running on Microsoft Windows for building digital twin models without the need for coding. This allows analysts and engineers without programming experience to quickly and easily create digital twins. Using this tool, you can build either a rules-based model, a machine learning (ML) model, or a combination of the two. The tool lets you first define a model and then test it prior to deploying it to the ScaleOut Digital Twins™ service. The tool prepares a zip file of the model that can used for deployment.

A rules-based model defines a set of numeric and string properties to be maintained by each digital twin instance and a set of rules which operate on these properties. Rules take the form of property assignments and IF-THEN tests, and they run whenever a message is received from each twin’s data source. They evaluate incoming message values and update the twin’s properties to maintain dynamic state information about the data source. They also can call built-in functions to perform computations, send messages, and generate alerts.

You also can create a no-code ML model that analyzes properties for spikes, trends, and anomalies using ML algorithms implemented by Microsoft’s ML.NET library. ML analysis also can be used to enhance a rules-based model. Please refer to the Machine Learning overview topic to learn more about the ML concepts employed by the ScaleOut Model Development Tool.

The following sections describe how to create and test rule-based and ML models from start to finish with examples.

• Prerequisites
• A Tour of the ScaleOut Model Development Tool
• Defining a Model
• Testing a Model
• Running simulations with multiple models
• Deploying a Model
• Sending Messages to a Deployed Model
• Model Samples