AutomatedBuildings.com Article - The Story Threads to Follow as Smart Buildings Enter a New Year

We’ve arrived at the end of another year, and the mystery of when adoption of Smart Building concepts will truly take off continues. The storyline that emerged in 2016 offered new clues and revealed certain dead-ends. Here is my plot summary of the who, what, where, how and why as we enter 2017.

Data Scientist

Who: Data Wranglers

The building controls industry is a central player on the data landscape today. The open-protocol community of the industry has welcomed bright technical talent from the ranks of HVAC family businesses, energy management start-ups, property management firms, building equipment corporations and other wandering paths. Our systems integration knowledge has expanded along with the evolution of web services and data transport methodologies. Our people know how to gather and structure building operational data for analytics, how to store it, and when to do batch versus real-time processing. That is data engineering and/or data wrangling according to this article from the Insight Data Science Fellows Program. But, we don’t often call it that. Could 2017 be the year we sync up a vocabulary with the rest of the data industry?

A great conversation about being welcoming and inclusive to a diversity of people –especially young people looking for a purposeful career–is available from the ControlTrends Network CTN. Download this podcast hosted by ‘Young Guns’ Rob Allen and Brad White as they interview guest Lindsay Baker, CEO of Comfy. They talk about how industry jargon can act like a wall keeping newcomers out. Certainly, the data world has its own impenetrable language. But data wrangler is a friendly, approachable term to use more in 2017.

What: Cyber-Physical Systems

As AutomatedBuildings Editor Ken Sinclair often explains the controls industry has been going about the business of digitalizing the physics of building operations ever since the first Direct Digital Control (DDC) devices were introduced as a replacement for pneumatics decades ago. But, in 2016 he noticed that the pace of that process was going from evolutionary to revolutionary. Almost every month of 2016 brought new announcements about ever more intelligent, powerful and versatile control devices for building systems. Edge Analytics Controllers (EACs) introduced in March by BASSG were among the first that he recognized as a brand-new category. “They have the brains of a smartphone and a chassis that can plug right into a control panel,” is how he describes them in this CTN podcast.

What to call the new category? He settled on “Maker Devices” in his October/November editorials. But, I think we will hear a lot more about Cyber-Physical Systems (CPS) in 2017. As you can read in the White House press announcement, the National Science Foundation’s Smart & Connected Communities initiative announced $4 million in new Cyber-Physical Systems (CPS) awards last September. The NSF defines Cyber-Physical Systems as:

“Engineered systems that are built from, and depend upon, the seamless integration of computational algorithms and physical components. Advances in CPS will enable capability, adaptability, scalability, resiliency, safety, security, and usability that will far exceed the simple embedded systems of today. CPS technology will transform the way people interact with engineered systems—just as the Internet has transformed the way people interact with information. New smart CPS will drive innovation and competition in sectors such as agriculture, energy, transportation, building design and automation, healthcare, and manufacturing.”

The controls industry has put the physical in cyber-physical for a long time. We won’t be cyber bullied. As Ken says in the podcast referenced above, “We’ve been supplying these boards for 30 years. We figured out what makes an input/output go on and off without the computer telling it too.”

Where: The Acephalous Network Edge

Acephalous means ‘without a head’ and has been used to describe societies without a permanent leader or chief. The computing/communications architectures of tomorrow are going to be non-centralized, distributed, self-organizing—i.e., acephalous, said a commenter in response to a new essay by Glen Allmendinger, President of Harbor Research. He was referring to this passage from The Failure of IoT Platforms.

Today, platforms for the Internet of Things are still a kludgy collection of yesterday’s technology and architectures that do not address the most basic development challenges. The world of smart communicating devices is mostly organized in hierarchies with smart user interface devices at the top and the dumb devices [often analog or serial sensors, and actuators] at the bottom. …As the Internet of Things opportunity matures, the sensor and actuator devices will all become smart themselves, and the connectivity between them (devices, for the most part, that have never been connected) will become more intelligent and the interactions more complex. In this evolving architecture, the network essentially flattens until the end-point devices are merely peers and a variety of applications reside on one or more [OT] computing devices.

“Hybrid Controllers that can act as either Global Controllers or Field-level DDCs are poised to change this situation. They will enable data flow that more closely mirrors the flow of air through a building. HVAC design practices will get back to the classic principles of easy discoverability and simple feedback loops in this next phase of Building Automation System (BAS) platforms. The impact is not limited to air control; Hybrid controllers will make the logic designs for controlling chilled water, hot water, demand response, integration of renewables and other building infrastructure networks more obvious, easier to maintain and more solid to build upon for years to come.”

Allmendinger’s essay also complements a new whitepaper from Harbor Research underwritten by SkyFoundry: The Future of Smart Systems and IoT Analytics. Both make the point that “the IoT platforms that have flooded the market are data traps and information islands.” They delve into why the client-server model of computing has served to prohibit true data interoperability and why this rigid hierarchical way of thinking needs to break down to enable the next era of machine learning to come into being.

Once again these points are consistent with those Alper Uzmezler, and I made just a little earlier with our November article, Today’s Smart Building Data Exhaust Maybe Tomorrow’s Machine Learning Gold. We also warned about the potential repeat of the Protocol Wars that have kept building equipment data in walled gardens—this time with cloud application vendors holding building operational data for ransom. The best strategy is to build on open-source software to the largest extent possible and to ensure that any cloud vendors collecting your real-time data and analyzing trends maintain useful software developers’ kits so that you can get the data when you need it.

Another guideline for building data wranglers that want to steer clear of data silos is to get educated about meta data strategies. As Scott Muench describes in his July article, The Strategy and Payoffs of Meta-Data Tagging a system like the Project Haystack methodology offers a way to add meaning to data that will be understood across the design-build-operate cycle and for years to come. Project Haystack brings the power of an open-source community to the data modeling challenge. It encompasses the combined experience of system integrators with long experience combining data from different sources and bringing it into value-added applications. In 2017, Project Haystack is holding its bi-annual conference, a not-to-miss event.

[an error occurred while processing this directive]Why: Machine Learning

In 2016 property industries continued to awaken to the fact that data is a valuable commodity and that buildings generate a lot of it. We are starting to see building owners and property management firms compete for tenants on the strength of their data platforms. Tenants want to lease space that supports the business-related digital services, personalized space and comfort, and energy efficiency goals they have in mind. This competition is leading the buildings industry in the same direction that manufacturing, transportation, retail and other industries have been headed for a while—toward machine learning and eventually artificial intelligence.

As Alper and I outlined in our September article When Will Machine Learning Reach Smart Buildings, “ML algorithms have advantageous self-correcting behaviors that will be the best navigators of a digitized world. But, these come at the price of being more complex to understand and work with than, for example, rule-based analytics programs. And they require a continuous and ample supply of structured data to deliver any meaningful results.” The property owning and management companies that, in 2017, put in place data strategies that will keep them in command of their data and help them get the most predictive value from it will be best positioned in the coming age of autonomous cars, smart cities, and the smart grid.

It feels like the large and small vendors and customers in the building operational data analytics software market understand where this is all heading. But size doesn’t lend the same advantages when delivering products and services to the acephalous network edge. (By the way, I don’t think acephalous will be a buzz word in 2017, or ever.)

Events	Want Ads
Our Sponsors	Resources