New Integration for Always-Ready Data Center Digital Twin Models

Garbage in, garbage out. A digital twin is only as good as the data you feed it.

This is especially true for digital twins of operational data centers. Taking a step back, a digital twin is a digital copy or model of an intended or real-world object or system. A data center digital twin is, therefore, a digital copy of a data center.

Data center digital twins, when powered by computational fluid dynamics (CFD), not only replicate current data center operations but also predict airflow and temperature distribution in the whitespace. In other words, CFD-powered data center digital twins can model how a data center will perform under any operating conditions. 

If the data being fed to the data center digital twin becomes outdated or doesn’t match its real-world counterpart, the predictions will in turn be outdated or irrelevant as well. This is a particularly concerning notion given that operational data centers are always changing.

If you’re concerned that maintaining a digital twin of an operational data center would require an impractical commitment of resources, fear not! At Cadence, we’ve come up with a rather simple solution to a complicated problem: integrations and automations.

Cadence’s data center digital twin solution automates model upkeep by integrating data feeds from popular operational tools, including environmental monitoring systems, data center infrastructure management (DCIM), and CSV spreadsheets. With these integrations, the digital twin model is always ready to predict with current data and minimal supervision.

And today we get to talk about one of the most exciting and important integrations yet: Honeywell’s IQVISION. This blog covers what the integration between IQVISION and Cadence Reality DC Insight looks like—what kind of information is shared and how this information improves the predictions the digital twin model makes.

Honeywell IQVISION and Cadence Data Center Digital Twins

IQVISION is a building monitoring and management solution or building management system (BMS) built upon the powerful Niagara 4 platform. IQVISION can integrate trend controllers, third-party devices, and internet protocols into a centralized software platform that is designed to manage buildings at an enterprise level. It can provide this function as centralized data logging, archiving, alarming, trending, scheduling, system-wide database management, and integration with enterprise software applications—all of which can be used for highlighting and investigating energy use within buildings. In addition, IQVISION provides a comprehensive, graphical engineering toolset for application development. HMTL5 support enables the customization of user interfaces that are viewable on diverse web-enabled computers, tablets, and phones.

While we go into more detail on our product pages, Cadence Reality DC Insight is a data center management software suite made up of two modules: Cadence Reality DC Asset Twin and Cadence Reality DC Digital Twin. The first is our solution for capacity and resource management, while the second offering builds upon the first with CFD simulation for data center optimization. The software suite is used to maximize uptime, capacity utilization, and energy efficiency.

The integration between Cadence data center software and IQVISION is a file transfer integration. It takes shape as a one-way data integration from IQVISION to Cadence Reality DC Insight. Specifically, three types of data flow from IQVISION to the Cadence platform:

• Cooling system setpoints
• IT power consumption data
• Environmental data

Cadence data center software captures the data from the BMS and stores it in a database. The software user can choose to store as much historical data as they require for further analysis. Once the BMS data is synched to the digital twin model, the software can visualize this data either inside the actual digital twin model or plotted on dashboards.

Let’s look at how the different data types from IQVISION improve the digital twin’s predictive capabilities and, in turn, data center operations.

Use Case: Cooling System Setpoints

As a digital replica of the physical facilities, the digital twin model also includes the simulation of the cooling system. The current setpoints are synced into the digital twin model to ensure they always reflect the current situation in the actual facility. Modeling the cooling system accurately is key to ensuring the digital twin model’s output is also accurate. Much of the cooling infrastructure is fixed, but setpoints can be adjusted as required by operations staff. The BMS is the source of truth for this data and syncing the digital twin model to the BMS ensures the model is up to date with any changes automatically, meaning that the digital twin model is always ready to predict cooling scenarios when needed.

Use Case: IT Power Consumption

The digital twin model also contains a full inventory of the data center’s IT equipment and a full model of the power network from utility to IT device. If this data is available from the BMS, Cadence data center software can monitor the power usage of different components over time and use this to set utilization on the IT load to enable the user to look at the data center in different states, such as the maximum power used over the last week or the average power used over the last day. Having an accurate view of the actual power usage in the data center is critical for making proper planning decisions. Connecting Cadence data center software to this data from the BMS puts this information at the fingertips of the people who need it most.

Use Case: Environmental Data

Environmental sensors track key data about the state of the data center. Information from these sensors—such as temperature, pressure, and humidity—is monitored over time and can be fed to the digital twin model. The data can be synced into a model to provide aggregated values for different objects. In turn, the aggregated data can be compared with the data from the digital twin simulation.

A key component of running a digital twin of an operational data center is ensuring the digital twin’s output matches measurement from the physical, real-world data center. If there is a deviation between outputs and measurements, that can be an indication that changes have been made in the real-world facility that have not been captured in the digital twin model, or that a device is not operating to spec and can be an early indicator of an issue. Either way, it represents a gap in the operator’s knowledge about a critical system and can be a starting point for investigation to alleviate the problem at hand.

TL;DR – This Integration Is Awesome

Seeing as we’re entering what many are calling “the era of AI,” the words integration and automation get thrown around a lot. As our lives get more and more digital in nature, we’ve come to expect everything to be fully integrated/automated/optimized—you name it—so at first hearing that we’ve established a new integration may feel like a given.

But for data center operations, this is groundbreaking.

For so long, CFD-powered digital twin models have been primarily used to design or retrofit data centers, but now, with this integration and others like it, we can do so much more. We can proactively move data from the physical data center to the model nearly effortlessly. We can predict faster, more accurately, and more consistently. All of this means we can make better, more informed decisions that will protect the longevity of our data centers.

This webinar does a great job of exploring what a fully integrated digital twin looks like in action. We’re also happy to talk about what a digital twin could look like in your own workflow—fill out our free trial form for a more personalized software demonstration.