Industrial IoT Platforms – IoT Performance Testing Best Practices

Blog • IIoT & Smart Connectivity

Every business demands growth and today’s digital transformations have given businesses ample opportunity for fast-paced expansion that was unthinkable just a few decades ago.

Internet of Things (IoT) advancements have unlocked new capabilities for just about every piece of technology including completely remote upgrades, instant monitoring and feedback, and system synchronization across the globe. With the ability to generate real-time data across an enterprise, IoT is revolutionizing the industry.

Companies are gaining access to data for their entire enterprise overnight, but converting that data into changes that improve scalability, optimize productivity, increase market share or generate profit can be an overwhelming challenge.

That’s why understanding how to manage high-volume industrial data and leverage purpose-built time series database technologies with your industrial IoT platforms is the key to preparing your company to be in the ideal position to take advantage of the opportunities, growth, and industry changes that lie ahead.

As the network of connected devices expands and real-time data flow surges, companies face the crucial task of selecting resilient, scalable, and agile Industrial IoT platforms. IoT performance testing plays a pivotal role in this process.

When exploring IoT and IIoT platforms, including time series databases and technical deployment strategies, it’s crucial to define your business architecture clearly. Start with a big-picture perspective to chart your digital transformation journey. Develop a roadmap that pinpoints IIoT connectivity needs, prioritizes complex data, and addresses potential challenges proactively, like malfunctions and blackouts.

But what happens as your business grows? Can your systems adapt to success? When one factory multiplies to ten, when your workforce expands from hundreds to thousands when local device networks go global, can your current setup keep up?

Imagine a company with multiple factories, each using separate systems and relational databases to monitor and interface. These systems share similar goals: remote asset monitoring and predictive maintenance.

At first glance, IoT performance testing could suggest no need for change – profit rolls in, and operations run smoothly. However, viewing each factory within the larger business architecture reveals redundancies. Operating independently might work short-term, but it escalates hardware expenses and complexity, consuming time, space, and money.

Transitioning to a unified IIoT platform designed for time series data simplifies IT management, cuts hardware costs, and expands scalability. It becomes evident that choosing the right industrial IoT platforms is essential for sustained success.

When beginning the journey to design or implement a unified industrial IoT platform for the massive volumes of time series data from sensors and systems, it is helpful to think of IIoT implementations as a three-dimensional scale problem; one that needs to undergo significant IoT performance testing.**

On one axis is the volume of time series data flowing in from connected devices. By understanding not only how many devices are expected to be connected but also how many different unique data items they’re expected to send and how often they will send them, you can start to understand how many writes per second (WPS) the solution architecture needs to support. This WPS calculation then factors into data processing requirements.

For every data point ingested, determine if any additional computation needs to be done immediately, or if that data processes in batches, perhaps on a scheduled interval. How complex are those computations? Do they only require the specific data item that was just ingested or does that data item need to be compared against previous data or against a known threshold? The number, frequency, and complexity of these types of data processing rules are part of the application logic load.

The last piece of the application logic load comes from the users who will be interacting with the system. How many users are expected to be using the system at a time? How complex are the operations to produce the user interfaces that they need to interact with? Is this use case a more real-time operational situation where users will be reviewing data as it updates and reacting to events and alerts? Or is it more analytical in nature where users will be running more complex queries with larger volumes of data to identify patterns and trends?

As you think about the load for processing time series data along these three axes, you can see where having a purpose-built time series database in your deployment architecture is a major benefit. Relational databases are, of course, capable of storing time series data, and they can do this job well at a smaller scale, but they are intentionally designed as a general data storage tool.

When choosing a time series data solution for your industrial IoT platforms, it’s important that the solution provider have the power and flexibility to meet all of your enterprise needs now and in the future.

PTC ThingWorx is the leading IIoT solutions platform for manufacturing, providing end-to-end capabilities for complete digital transformation. ThingWorx typically provides two different types of persistence providers or database storage locations. The model provider is always a relational database and stores general system data as well as metadata about the model being used for the application. However, one or more data providers are responsible for the runtime data ingested against that model, elements such as value streams from connected devices or from other integrated external systems. This data is almost always time series data.

While PTC ThingWorx customers can build custom integrations with almost any database, InfluxDB is the only time series database that is natively integrated into ThingWorx out of the box. With its powerful processing capabilities, extensible programming language, easy-to-build templates, and hundreds of open-source Telegraf data collection plugins, using InfluxDB as a key part of your PTC ThingWorx implementation enables you to optimize performance to a degree that no other time series database is able to approach.

In fact, InfluxDB is so trusted that it is embedded into PTC Cloud, making it the de facto standard for high-scale PTC ThingWorx implementations.

The features that make InfluxDB so trusted by this IIoT giant are useful at every stage in a business’s lifecycle. Utilizing InfluxDB early in a company’s digitization not only prepares businesses for future scale but can also reduce hardware costs, taking advantage of the high-performance data ingestion and crew performance capabilities that are provided by a time series database.

Using InfluxDB as the time series database with PTC ThingWorx optimizes industrial IoT platforms with data collection and storage while using about half of the total hardware of the original database system. Even at a small scale, using InfluxDB as the time series database is more efficient, effective, and profitable than other options.

The implementation that included an InfluxDB time series database was also able to handle almost twice as many service requests and about four times as much data ingestion as the comparable relational database configuration with a comparable amount of hardware.

Ultimately when undergoing IoT performance testing, the PTC ThingWorx platform paired with InfluxDB allows for long-term business benefits including more growth with less hardware, faster service and data processing, and an overall more efficient net enterprise result, which all translates to more potential for profit and market share growth.

No matter where you are in your digital industrial journey, you will need the solutions and systems that provide the most benefit now and will provide the most flexibility for future growth.

Selecting PTC’s ThingWorx with InfluxDB allows businesses of all sizes to implement, manage and utilize IIoT data to its fullest, removing the obstacles that can impede a company’s ability to achieve high growth.

Learn more about using InfluxDB with PTC ThingWorx.

TriStar PLM Solutions welcomes questions. Feel free to CONTACT US if you can’t find what you’re looking for, or call us at 800-800-1714