Industrial IoT applications: four DIY enablement features

In today’s manufacturing ecosystem, production data analysis plays an ever-increasing role in measuring performance. For many industrials, access to clean, contextualized production data feels like a pipe dream. Yet for others, especially those with advanced maturity, easy-to-use IIoT ecosystems are at the front and center of their operations. 

Regardless of where a company falls along this data spectrum, manufacturers must equip themselves with technological solutions that allow data to be extracted from the production chain. However, data extraction isn’t the end goal: companies must know what to do with the extracted data. 

Modern industrial data manipulation tools can require a lot of technical skill to use, and data analysis is often tackled in a different program (e.g. Excel, MiniTab, or Tableau). These tools offer user agility but they also have their limitations. As a result, manufacturers must use data management strategies that are sustainable, industry-specific, and easy to scale.

Industrial IoT (IIoT) applications, whether they are built or purchased, can provide manufacturers on both ends of the data spectrum with new opportunities. Applications make it easier to scale discoveries while also providing a larger number of users with the capability to analyze data. 

Building apps within IIoT

Whether manufacturers are getting started with Edge or Cloud infrastructures or have already implemented automation layers, there are new expectations for Industrial IoT (IIoT) platforms to provide Software Development Kits (SDKs). 

SDKs are a developer’s playground. They are designed to fast-track the development of applications through different software platforms. While many deem SDKs as a prerequisite for an IIoT Platform, an SDK alone won’t solve some of the deeper challenges with industrial data. 

When creating industrial IoT applications, it’s important to consider the necessary degree of interoperability for a given application. For example, SaaS platforms that leverage ​​containers add a new layer of interoperability as they contain all of the necessary elements to run in any environment. 

The most effective solution to obtain sustainable gains and prepare for the future towards modern and efficient data processing solutions is to use a specialized platform for citizen developers. 

Braincube, an IIoT platform, provides a ready-to-use SDK ecosystem (including widgets and APIs) in addition to off-the-shelf applications. Braincube’s platform suite offers a set of advanced industrial data processing components. Specific connectors are accessible through a low-code development environment open to all users. A no-code toolkit of graphic elements allows teams to fast-track the construction of advanced, easily-shareable dashboards and other data visualization apps.

In short, an SDK makes it efficient for IT or other data teams to rapidly develop and publish new, secured industrial IoT applications and components. Braincube’s unique platform, paired with its SDK, provides a common environment for addressing both citizen developers and IT developers in a single space.

But what does this balance of build and buy look like in practice? Here are four ways that Braincube equips manufacturers to build their own IIoT apps:

Four approaches to building industrial IoT applications within Braincube’s IIoT Platform

Building apps from Node-RED subflows

Node-RED is a graphical low-code programming tool for wiring together hardware devices, data processing, APIs, and online services in new and interesting ways. Leveraging browser-based flow editing, it is compatible to extend data for a wide variety of use cases including manufacturing. Node-RED essentially allows data flows to be configured. 

Braincube’s low-code environment benefits from Node-RED’s powerful solution and rich ecosystem of modules. However, Braincube builds upon Node-RED’s foundation by customizing it for the industrial environment. 

Within Braincube, manufacturing employees can easily drag-and-drop pre-existing nodes into place based on their flow of data. This allows citizen developers to visually build the acquisition, processing, and dissemination of production data. This is done through the use of specialized nodes which are drag-and-drop and assembled to each other by links representing the flow of data.

Teams can also build their own applications using a subflow within Node-RED. A subflow is a consolidated group of individual nodes that are seen as a singular entity. 

For example, say you want to receive a notification every time your data flow stops. You can build a subflow that, when triggered, sends an automated message. This can be followed by a triggered reminder (i.e. second message) after a certain time frame unless extended or reset by the recipient. This subflow would query and notify a specific user when there are any problems in their data flows.

Separately, these are different nodes and wires. But when grouped together, they create a streamlined subflow that supports your procedures more easily. 

Once a subflow is created in Node-RED, it can be published as an app and distributed throughout your organization. The subflow described above is a pre-built Braincube industrial IoT application known as Timeout. From here, the application can either stay within the Node-RED library or can be used directly in the Braincube IIoT Platform. 

Braincube has also supplemented the open-source Node-RED suite with ​​enhanced security, easier scalability, fleet and flow management, and additional nodes customized to the specific needs of manufacturers. 

As data diversity continues to expand, having no-code/low-code capabilities for data-flow management and IIoT app development is critical for citizen data scientists and IT teams.

(Still trying to decide whether to build or buy applications? Check out our overview.)

Create app templates in Studio

Braincube’s Studio App offers a low-code/no-code environment to combine real-time and/or historical production data from any IT, OT, and third-party system. In short, it is a visual Braincube data flows manager. Studio is equipped with a WYSIWYG builder, making it possible for users of all skill levels to create data visualization apps in a few clicks. 

For example, let’s say your team needs to monitor carbon output. Studio’s pre-built widgets can be customized to pull the necessary Cloud, Edge, or other data source (which has been configured using Node-RED) into a highly personalized application that’s always up to date. Depending on your data, the application could display energy, gas, and water consumption during production. It may extend beyond production-specific metrics to cover the fleet used to deliver end products or the emissions used for employees to come to work.

Regardless of what metrics you track, Studio allows hyper-personalization. You can choose what data is displayed, how data is moved across the organization, and even the exact fonts or colors used in the app palette. Ultimately, manufacturers have clear visibility into the most important up-to-date information, whether it is a single machine’s status or a complex company-wide KPI like carbon footprint. 

With endless customization within Studio, you can insert ready-to-use apps, add reports from select integrated apps, or build a custom app using our open API or WYSIWYG builder. These app templates can be shared and reused across the organization, improving collaboration and cross-department discoveries. 

Program your own Industrial IoT applications

Braincube helps users program their own Industrial IoT applications in two ways: facilitating app development or enabling users to bring in custom apps.

For those looking to build their own apps, Braincube also allows users to develop apps directly in Studio, which sits within the IIoT platform. Users with deeper technical skills, such as Python or java coding, can also leverage the power of the platform. These users can write apps outside of the Braincube platform and then add them to the app manager for private or public use. 

Just like with any other software, users can open the editor of their choice, write the code, and save it. Open up the terminal provided by your OS and locate the saved file. Programming in Python or Java allows users to leverage APIs so that they don’t have to reinvent the wheel. 

For example, you may want to use a Translator API for the backend coding and a graphical user interface for the front end. These tools can make getting a usable app to end users simple. They also improve your capacity as a manufacturer. In both scenarios, apps can be executed inside or outside the platform based on a user’s choice.

Run your own data science models

​Braincube’s APIs, SDK, and container-based deployment workflow allow customers to deploy custom-made apps and models.

“Braincube enables us to perform different calculations,” said a lead engineer at Kimberly-Clark. “It allows us to run custom Python models to implement all of these Machine Learning algorithms. Being able to generate a lot of these very high-quality predictions using just a minimal sample size is quite extraordinary.” 

In addition, Kimberly-Clark leverages the Braincube API to pull data and models (including offline models) directly from the cloud. 

Through Braincube’s Python connector, data scientists can integrate Braincube data seamlessly with Python. Users can leverage bi-directional APIs to feed Braincube’s centralized, clean, and prepared data from their Digital Twin into Python. This enables teams to utilize custom functionality.

When considering industrial IoT applications and platforms, strong consideration should be placed on security and interoperability. For example, Docker is an open platform for developing, shipping, and running applications; this is the platform Braincube utilizes. Docker “provides the ability to package and run an application in a loosely isolated environment called a container.” By using the Docker platform, apps are quick to deploy and can be treated as microservices for better speed and improved security protocols. 

Docker containers can run in a mixture of environments. This includes a VM, laptop, private cloud, and more. Docker is commonly used in manufacturing already, given the monster amount of data collected by manufacturers. Within Braincube, each model tagged as having production value is automatically built and packaged in a self-contained Docker image. From here, it’s ready to be operationalized on production data flows.

Summary

Some manufacturers prefer to build solutions in-house, including industrial IoT applications. However, just because an organization prefers to do something in-house doesn’t mean there aren’t benefits to working with a third-party company.

Solutions like Braincube are designed to be flexible and adapt to the needs of a variety of users. Plug-and-play applications can run alongside custom in-house applications. Braincube also offers a variety of enablement features to enhance the building and distribution of industrial IoT applications that are built in-house.

As your organization moves forward, it’s important to bring in a wide array of tools for continued growth. To learn more about how Braincube can partner with you on your journey, contact us to receive a personalized demo.