The Road to Transparency in a World With Billions of IIoT Devices

James Lee, CEO & President, Cimetrics Inc.

Last month, I wrote about how analytics creates transparency and the relevance of this for players in the building automation industry. The bottom line is that low-hanging fruit is there for businesses trying to make sense of the analytics trend sweeping the industry. If you had not read the piece, it would be helpful to do so to provide a context of this month’s perspectives (the article on this link).

Now we know that analytics is a real destination, we can focus on how to get to there, which is where IIoT comes in, the Commercial & Industrial big brother of the consumer IoT that is slowly creeping into our daily lives.

IIoT, IoT, and their elder cousin M2M have been around for a while now, since 2004 when many attended the inaugural M2M Expo in Dallas, TX. For the decade since, M2M/IoT have mainly been abstract concepts, a Jetsons futuristic world of talking home appliances and flying cars. Well, my friends, this futuristic world is emerging around us, “Alexa, please order laundry detergent.”

What makes IIoT different from IoT?

There is one major way that IIoT is fundamentally different to the IoT. Since IoT is for consumers, the aim is to make everything as much as possible be plug-and-play. You buy a new IoT-enabled bathroom scale; you bring it home, it connects to your WiFi, it contacts the cloud service that manages your health and voilà, your weight is in the cloud. In IoT, there is little configuration, and minimal dependence between devices, they are mostly discrete. In other words, if an IoT-enabled toaster breaks, it’s not likely to affect the operation of the bathroom scale.

IIoT systems in buildings, on the other hand, are engineered systems. Each piece of equipment is configured, installed, programmed to connect to and work in a specific building, for a particular function with a unique set of other equipment in that building. Building systems are created by humans, where there is ample opportunity for installation errors, lack of training, manufacturing or supply variances and misaligned business interests. There are also changes in the needs of the facility, and the natural law of entropy to mess things up (leaks, components breaking, low lubrication, wear, and tear, etc.).

Of course, the primary purpose of a system is to run the equipment in the building. But a key to the IIoT value proposition is the ability to keep the system working at peak performance for the building occupants over the life of the building. You see, IIoT in buildings is a constantly evolving beast. The system at initial commissioning is not the same as it is a few months and years hence. Any system trying to analyze such a building needs to adapt with little or no human interaction. Such a system has to model the workings of HVAC and related equipment, understand their rules of operation and determine the cause of any issues.

So, what makes up IIoT in buildings?

While IIoT is often discussed as an amorphous creature, it’s not hard to break it down to key components, especially in the context of buildings. This view is important since most of these elements are quite easy for building automation experts to grasp and leverage.

• Devices: These are the basic building blocks of any building automation system: supervisory controllers, thermostats, unitary controllers, actuators, etc. We’ve found that a key to proceed in the IIoT world is to ensure the uniformity of the data points on these devices, points that make up the digital fingerprint of these systems. The best way to ensure consistency is to ensure that these devices are native BACnet, which also make them inherently discoverable and interoperable. Also, BACnet provides a great deal of metadata to describe a device’s function and context, which, in turn, enables IIoT to create a digital model of the physical components of the building system. Metadata had always been an important part of BACnet from its inception nearly 30 years ago, and current work by the BACnet community continue to strengthen this key attribute. A recent addition to the BACnet standard is the ability to attach descriptive tags to any BACnet object, and an initial set of standard tags is forthcoming.

• Networks: There are two parts of networks in the buildings IIoT world: inside and outside of the building. Since we believe that the connected devices should be BACnet, the in-building network should, of course, be BACnet, which if well designed will remove the need for any proprietary gateways. It’s worth noting that gateways aren't just unnecessary, they can destroy or diminish much of the value of BACnet devices by stripping out metadata, making it harder for IIoT analytics to be performed. Outside the building, there, of course, has to be a secure Internet connection. Note that it’s unnecessary to expose everything that happens in the building, nor does it need to be real-time. We’ve found that installing a BACnet-aware cache-poller (a data collection, buffering and communication device) provides a balanced solution to ensure security, access to necessary information and efficient use of communication bandwidth.
  

• Cloud: Currently one of the hottest trends in Tech and IT, the cloud adds a great deal to IIoT analytics of buildings. For us, it’s best to think about the cloud as a digital twin of the physical building, complete with all relevant metadata, structure, and context. This approach digitally recreates all of the important relationships between devices, configurations, control strategies, dependencies as well as current and historical data necessary to continuously check the effectiveness of the system. A model-based approach enables us to look at the system with knowledge of how it is supposed to be working, and identify behavior not consistent with nominal performance. This approach is far superior to statistical analysis and expert system approaches that are mostly ignorant of the broader context of the engineered system. Additionally, the cloud provides data for use by us humans for deeper analysis, reporting, comparison, etc.

How to tame this beast!

There is one aspect of IIoT that is worthy of additional discussion; the size and scope of the problem. When BAS professionals think of points, it’s quite natural to think about specific points of some VAV or other devices. There may be a few, a dozen or more points of relevance, and that is a very human scale. You can write them down, enter into a spreadsheet and otherwise manipulate them manually.

When thinking about IIoT and the analytics necessary to perform in the cloud, the number of points dramatically increase to hundreds, thousands, and with massive facilities, it can be in the millions! At these quantities, it ceases to be a human-manageable set of data; it’s simply impractical to try and manage such a dataset in any traditional way.

Watch Maya’s Mickey McManus on the tyranny of numbers https://goo.gl/Ovyqvp

To fully grasp the business opportunities with IIoT in buildings, it is critical that you understand that it is not possible to think about large-scale continuous and automatic analytics the same way as thinking about the design, installation, and maintenance of BAS devices and systems.

An entirely new approach is required for this task, as outlined above. Thankfully, the BAS industry is well prepared with nearly three decades of development, maturity and now mass adoption of the BACnet standard. It’s interesting to see the long ASHRAE standardization work now finally paying dividends in the new world of IIoT.

We at Cimetrics believe that the formula for this is quite simple.

Native BACnet + IIoT Models based Analytics = Business Opportunities