Navigating the Platform

Finding your way around the platform

The Rayven Platform consists of Rayven Live, the interface for viewing dashboards and user-accessible widgets, and Rayven Workspace, the backend used to create workflows.

Rayven Workspace addresses all your software needs in a simple-to-configure platform with enterprise-grade security and back-ups.

Architecture and data flow

When using Rayven to create a solution, the key areas are the Workspace, the Workflow Builder, and the Machine Learning Workbench.

Rayven workspace: a simple-to-configure all-in-one platform

The Rayven workspace is your initial entry point to the Rayven platform. You can use it to build and deploy your own solutions or select a pre-built solution for a specific vertical or use case. The workspace is also where you can add and manage users and devices.

Device Management

All devices and assets can be registered or deregistered from Device Manager in the Rayven Workspace. You can do this individually, incrementally, or by bulk upload.

The device record contains the device name and ID. It also holds the device's IP address and latitude and longitude, if these are specified.

Devices are registered in device groups and maintain hierarchy metadata like parent-child relationships, time zone, location, and icon(s). You can organize hierarchies into parent-child structures, for example, machine/line/factory or unit/floor/building/postcode/state.

All devices are monitored for their last communication time and display whether they have been active in the preceding 24 hours. The device inventory can be enumerated and queried by group, in bulk, or individually.

User Management

Every user in Rayven needs to be associated with a user group. User groups allow you to easily control what data and dashboards each user can access when viewing Rayven Live.

The Rayven Workflow Builder

Our Workflow Business Logic Modeller is a drag-and-drop interface that enables you to easily create rules and connect systems without writing a line of code. You can use it to combine different data sources, perform complex calculations, add AI, and visualize the results.

Connective nodes enable data inputs, processing, and visualizations. The Rayven platform has over 60 out-of-the-box nodes. For more information on configuring nodes and workflows, see Section 3.2: Workflows.

Data ingestion, consolidation, and transformation

Inputs

The Rayven platform includes a wide range of built-in nodes which enable ingestion of device, asset, and system data at customizable frequencies.

Your input options include dynamic API nodes that enable you to connect to third-party systems without coding.

Given Rayven’s modular approach, we can add new nodes to connect and ingest other data sources if your business requires it. In addition, for more complex system integrations, Rayven can create custom ingestion services.

Functions

Use 'functions' to manipulate data using custom logic, formulae, or JavaScript code.

Logic and processing nodes can convert ingested data from industrial protocols to human-readable form, perform complex calculations, and manipulate it with conditional logic.

Tools

At any stage of the workflow, the debug nodes under ‘Tools’ can be used to query the data.

Live Dashboards & Analytics

Processed data can be pushed to dashboards using the visualization, control, and output nodes:

• Visualizations can push data into charts, tables, and maps.
• Controls enable you to make the dashboard interactive using buttons, dropdown menus, and text fields.
• Outputs can send data to third-party systems or reports.

Rayven’s dashboards are fully white labelled, customizable, and mobile-friendly.

You can monitor the health of devices using Rayven’s dashboards and tables, as in the device inventory table below:

Machine learning and predictive analytics

Selecting ‘Machine Learning’ on the Workflow Builder allows you to create a repository and send data to the Rayven Machine Learning Workbench. Use this workbench to create your own Machine Learning models or import any existing Python-based model. Train and test these using your data, then deploy them into your IoT monitoring and management solution using Rayven AI Dynamix.

AI Dynamix is a modeler and engine programmable by anyone using drag-and-drop logic. Use it to detect anomalies, predict failure, and forecast output.

To create a machine learning model:

1. Select data of interest from the workflow builder and prepare it for analysis. 
2. Choose a model to apply. Rayven offers pre-built models as standard, or you can drop a Python model into the machine learning flow.
3. Train and evaluate your model using automated settings, or tune the hyperparameters based on your requirements. 
4. Add as many models into the flow as desired and train them with the same prepared data.
5. Evaluate and compare the results to select the best model for your use case.

Once you save a model, it is immediately available for deployment in the workflow builder. You can combine it with other data for further processing or display it in a visualization node.

The platform allows you to run multiple models simultaneously and compare outcomes for continuous improvement.

Reporting

Most reporting use-cases can be delivered out-of-the-box, even for non-real-time data. Rayven supports the following reporting methods:

• Rayven’s reporting dashboards, with full-page filtering for historical charts.
• Downloading and printing PDF reports from the real-time interfaces.
• Triggering reports based on logic or an event.
• Scheduling email reports with chart widgets as attachments.
• Imbedding chart widgets in the body of email reports.
• Attaching reports as CSV files within emails.
  

Lets get started with a 5 minute introduction to the Rayven platform

In this video, you will:

• Get an overview of the Rayven platform,
• Learn how to navigate the Rayven Manager,
• Learn how to change basic settings,
• Gain an understanding of widgets, device labels and Rayven’s AI Dynamix machine learning engine,
• Discover the fundamentals of solution building.