As pioneers of the Industrial DataOps space since 2018, HighByte and its users have enabled interoperability and digital transformation for the world’s leading industrial organizations.
As the Intelligence Hub is broadly adopted in an organization, we often see an ever-expanding backlog of use cases and a desire to involve a more diverse set of users to support their delivery. This elevates the importance of ease-of-use, observability, and governance. Champions of the Intelligence Hub must continue solving interoperability problems without compromising on scope or pace to support organization-wide adoption.
Based on my experience, HighByte Intelligence Hub users want to:
- Connect deeper into key systems in their architecture
- Work with data structures more capably and flexibly
- Hydrate these data structures with more control and efficiency
- Organize this data semantically for not just one, but multiple target systems and access patterns
- Design, govern, and augment their unified namespace (UNS) architecture beyond the traditional scope of publish/subscribe messaging technology and across technology stacks.
For industrial companies, Intelligence Hub version 4.0 breaks down the dichotomy between the advanced integration capabilities their digital transformation initiatives require and the ease-of-use their growing teams desire.
It starts with a completely new UX and centering configuration around Pipelines with its intuitive graphical builder. In version 4.0, Pipelines replaces Flows to become the future of data movement within the Intelligence Hub. The latest release also introduces new modeling functionality that allows models to be created with hierarchy and composition. This streamlines model definitions and configuration of complex data structures.
HighByte Intelligence Hub version 4.0 ushers in the next generation of Industrial DataOps. Version 4.0 delivers a new configuration paradigm centered around Pipelines. It focuses on simplifying the user experience, data modeling, and advancing the UNS architectural pattern with all-new Namespaces and Smart Query functionality, and it introduces a new native integration into Inductive Automation’s Ignition platform. This post will explore each of these new capabilities in more depth.
New UX + Pipeline-Centric Configuration
Since its introduction in version 3.0, Pipelines have demonstrated their ability to sequentially read, shape, and publish industrial data in a powerful, intuitive, and observable way. Pipelines have proven invaluable in progressively engineering optimal, contextual payload data with top-notch monitoring and troubleshooting.
Users of HighByte Intelligence Hub version 4.0 will notice a new UX across the product, providing a cleaner digital workspace. Several core features have received a major refresh, including Connections, Models, and Pipelines.
The Pipelines UX has been refreshed with a left-to-right orientation of stages, an adjustable panel for stage configuration, and mega menus for Stage, Replay, State Variables, Usage, and Test features. Thoughtfully accounting for all the new functionality within Pipelines, this maximizes viewable pipeline stage content while viewing, configuring, and monitoring a pipeline. This is ideal for larger and wide-format screens. Additionally, each Pipelines stage now includes callouts indicating state and execution count, allowing users to see the health and activity of a pipeline at a glance.
Supporting a key objective of version 4.0, Pipelines have evolved to make the Intelligence Hub even easier for a broader set of users. In version 4.0, configuration is centered around Pipelines, which are now at the core of all solutions built in the Intelligence Hub. Users can visualize the flow of data and how configuration objects interact end-to-end.
As mentioned in the introduction, Flows are now a subset of functionality within Pipelines. Pipelines are self-triggering, driven by three new stages: Event, Polled, and Flow. Whether data is moved on an event or interval basis, the exchange and remediation of data between sources (i.e., Inputs, Conditions, Instances, Namespaces) and targets (i.e., Outputs) now occurs in one place. Pipelines allow users to interrogate and observe not just configuration, but what is happening to their data at every step.
Next-Generation Modeling
Modeling has always been a cornerstone and differentiator of the Intelligence Hub. It allows industrial organizations to bring order to the often-chaotic landscape of industrial data sources. It drives the standardization and scale of how data is delivered and consumed. As users have recognized the value of the Intelligence Hub as a dedicated modeling solution for their industrial data, their use cases and data modeling requirements have continued to evolve. HighByte Intelligence Hub version 4.0 elevates modeling with more capability around defining and managing complex data structures, re-using and composing data models, as well as flexibility in hydrating data models.
Models: UX, Composition & Structure
The Models and Instances user interfaces have been fully revamped, featuring a visually intuitive attribute tree. From a single glance, the tree indicates the “shape” of a model and presents options to alter it. The tree also indicates whether an attribute is a simple data point, a nested model, or complex object with child attributes. This complements the new capabilities surrounding model composition and structure. Model attributes can now be nested to create depth or complex objects within a payload. Models can reference other models to extend their definition. Additionally, a single Instance can now be derived from multiple models. Together, these enhancements simplify configuration and make defining complex data structures even easier. This functionality is particularly useful for enforcing data standards on integrations involving data lakes and modern SaaS applications with sophisticated API endpoints.
Instances: Initialization & Processing Improvements
Under the hood, Instances have been streamlined to process attributes and expressions more efficiently. We’ve also added explicit options to control when and where JavaScript expressions are employed. Users can elect to transform data at an individual attribute level or—with the new Initialization feature—as a single batch process. Initialization, as the name implies, is a JavaScript expression that is evaluated first within Instances. It provides the capability to pre-process input reads and data transformations ahead of Instance attributes. It is ideal for sharing or sequencing dependent inputs and data transformations. For example, in a single step, one could tie together the inputs of a MES, OPC server, and CMMS as a sequenced lookup, transform the results, and then associate with Instance attributes. There are many use cases where the value of an attribute may be dependent on another. Initialization simplifies these scenarios.
With version 4.0, Instances provide granular control over data processing. From tuning for performance to sophisticated data extraction, users have the flexibility they need to put their data models to use.
Modeling in Pipelines
Making their first appearance within Pipelines, the Model and Model Validation stages can model incoming event payloads within a Pipeline and validate them against a model definition. Complementing the Intelligence Hub’s explicit modeling capabilities to form data structures through Instances, these new Pipeline features enable users to implicitly enforce modeling standards on already-structured data with ease.
For example, suppose 500 robots produce a similar data structure for their servo motors. Instead of explicitly modeling every single data point for each joint’s servo motor for each robot, one could validate that the incoming data structure meets modeling standards and then remediate or reshape the data structure as necessary.
From sophisticated model management and definitions to granular data processing to explicit and implicit model enforcement, HighByte Intelligence Hub version 4.0 delivers the next generation of industrial data modeling.
Namespaces + Smart Query
The Intelligence Hub is used to both orchestrate the shape and delivery of individual payloads (or datasets) of industrial data and to organize them. At scale, this orchestration is imperative given that the semantic relationships between datasets can be just as valuable as the datasets themselves.
To solve for this, version 4.0 introduces Namespaces as a new, primary component of the Intelligence Hub for organizing datasets and their relationships. Namespaces create a well-defined JSON document representing the factory or what is connected to the Intelligence Hub. The document contains the organizational hierarchy of the business as well as inputs and modeled datasets with metadata. These datasets could be anything, from the state of an asset to an asset’s pump condition to the results from quality inspections. As the namespace is being configured, its contents can be previewed as a JSON payload in whole or partially by node.
Intelligence Hub Namespaces provides a visual approach to the design and governance of a UNS—or really any hierarchal organization of datasets. Users can design the namespace with an intuitive, graphical node-tree to create nodes and organize their hierarchy. Alternatively, users can import the namespace definition from external sources using CSV or JSON formats.
Accompanying Namespaces is a new Pipelines stage called Smart Query. This feature can intelligently extract the organized and modeled industrial data within Namespaces into a pipeline. It can seamlessly query data by asset hierarchy, model type, attribute value, and more. For example, Smart Query can retrieve only robot data specifically from an area within a plant across many lines with multiple asset types that are actively running. And with Pipeline processing, this data can be packaged for delivery to an MQTT broker, historian, data lake, and relational database with consistent semantic hierarchy.
Ignition Module
Intelligence Hub version 4.0 introduces a new native integration with Inductive Automation’s Ignition platform. The integration consists of a HighByte-developed Ignition Module, deployed in Ignition Gateways, to which the Intelligence Hub securely connects.
The module provides direct, performant access to Ignition’s internal Tag system. It can browse and read Ignition tag providers including their folder, tag, and user defined type (UDT) contents. The connector can also write to tag and UDT values within a tag provider. With the Create option enabled, the connector can even create tag providers (as well as all folder, tag, and UDT contents) in hierarchal form.
The Ignition Module unlocks new capabilities beyond existing approaches such as OPC UA or MQTT/Sparkplug. Whether Ignition is deployed as a SCADA or an IIoT platform, the Intelligence Hub’s integration provides myriad benefits.
Wrap Up
Scale—namely integrating more with less resources—is a core tenant and value proposition of the Intelligence Hub. The next-generation Industrial DataOps capabilities available now in HighByte Intelligence Hub version 4.0 enable industrial organizations to tackle more interoperability problems of greater complexity to achieve digital transformation outcomes.
Beyond what you’ve read about here, many new features and capabilities are included in the latest release. To learn more, please check out these additional resources:
- Read the product announcement
- Explore the release notes for details on all new features and fixes.
- Request a free trial or log in to your existing account to test and deploy the software in your unique environment.
