IOT: How to make buildings more energy-efficient

IOT: How to make buildings more energy-efficient

  • VIDEO: Above, San Francisco CIO Miguel Gamino opens the IoT World Congress in Silicon Valley.


Moderator Chua.

Moderator Chua.

by GODFREY CHUA, Principal Analyst, Machina Research | June 23, 2016

Connected solutions deployed in buildings represent one of the most meaningful opportunities in the IoT realm. In fact, according to Machina Research, use cases revolving around building automation, security and smart metering will account for nearly 6.8 billion connected devices by 2020—or over a third of total all machine-based connections.

Last month at the Internet of Things (IoT) World event in Silicon Valley, Santa Clara CA, I moderated a panel discussion of top industry experts to discuss how IoT will help us move towards improved energy efficiency in buildings. Our panel: Dave Bartlett, CTO of Current, powered by GE; D. Young Nguyen, Senior Manager, Metering Services & Engineering, PG&E; Matt Gates, Director of Intelligent Services Offers, Trane; and Matt Oberdorfer, CTO, Frost Data Capital.


Godfrey Chua (GC): What do you see as being among the most important priorities for building managers/owners with regards to energy efficiency? In your experience, where is the greatest amount of “waste” occurring (effectively, how bad is the problem)?

IBM vet Bartlett.

IBM vet Bartlett.

Dave Bartlett, Current (DB): I am pleased to see the progress that many building managers/owners have made by implementing advanced EMS [energy management systems]. We have been able to tackle better energy management and asset management, bringing better automation and making building staff more productive. 

With the advent of the Internet of Things, however, we have the opportunity to move to a new generation of building management—one that is enabled by a more pervasive set of sensors to make the built environment more contextually adaptive (i.e. our buildings become task-aware, with the ability to adjust in real time). This “Gen 2” play, as some are calling it, will enable us to move from making our facilities engineers more productive to making the building occupants more productive.  

we have the opportunity to move to a new generation of building management—one enabled by a more pervasive set of sensors to make the built environment more contextually adaptive
— Dave Bartlett, CTO, Current

One area of waste I see comes from concurrent heat/cooling situations, even in many of our LEED-certified buildings. An example is occupants complaining about “freezing” on the north side of modern air-conditioned buildings.  Many are plugging in heating units under desks in the middle of summer without the knowledge of building /energy managers, exacerbating issues.

Frost Data's Oberdorfer.

Frost Data's Oberdorfer.

Matt Oberdorfer, Frost Data (MO): The priorities for energy managers are clear: reducing operating costs and energy consumption by addressing those heat/cooling situations. The challenge is that many of them don’t even know how much energy is wasted and where. And here is why... According to the Forbes Insights report, No More Wasted Energy, there are nearly 5 million industrial and commercial buildings in the US and only 14,000 energy managers—one for every 360 buildings. In addition, those few energy managers lack visibility and awareness because of ineffective, disparate sensors and the lack of analytics apps that provide actionable insights. 

How big is the problem? Industrial enterprises and organizations with $10 billion of annual revenue typically spend somewhere between 3% and 5% of their revenue on energy, yet most senior executives remain unaware of this. That’s $500M for one company. So, all commercial and industrial entities spend $60 billion in unnecessary energy annually. That’s how big the problem is.

Young Nguyen, PG&E (YN): The energy efficiency is often a reflection of the customer’s/building owner’s affordability.  Where PG&E can help customer to save energy by offering energy efficiency programs, such as air conditioner curtailment incentives, LED lighting, refrigerator rebates, home insulation, etc. The waste of energy is often related to space that is not being used at the time as opposed to a location of need.  It may be as simple as the lights being on or the heat in an empty room.

Trane's Gates.

Trane's Gates.

Matt Gates, Trane (MG): Energy efficiency is not a "one time" event—you have to continuously monitor and react. Waste is occurring everywhere, which is why you need to know where it is coming from, but also realize it is doesn't always comes from the same places. Buildings are dynamic and are always changing—what worked yesterday to operate the building may change tomorrow as the building is being used.  That said, typically the largest opportunity for energy efficiency is to continuously manage occupancy schedules.  Our experiences are that energy savings from five to 30 percent are possible in most buildings by just being diligent with keeping occupancy schedules accurate.

GC: How well aware are asset owners of the challenges to energy efficiency? Is there a sense of urgency in looking to address building energy efficiency?

DB:  I think we need a greater sense of urgency, and that starts with awareness and transparency of issues up to and including building owners. IoT can help with that.  For example, the pervasive set of data that we can now stream from GE LED fixtures equipped with sensors—coupled with the set of data we already have—can provide new analytic and visualization opportunities that can clearly show the opportunity for efficiency, and thereforeinstill a new level of urgency.

MO: A common response is “Let’s just build more eco-friendly, energy-efficient buildings.” But that’s not solving the problem. Why? Because approximately 85% of existing commercial buildings and industrial facilities in the U.S. will still be standing in 2030, more energy-efficient construction cannot be the sole solution to cope with the energy waste challenge. We believe that emerging IoT analytics solutions will make the need for energy-efficient buildings very visible, and urgency will follow. That’s why we are actively looking into technologies and business models to invest in that help accelerate this trend.

Emerging IoT analytics solutions will make the need for energy-efficient buildings very visible, and urgency will follow
— Matt Oberdorfer, CTO, Frost Data Capital

YN: In general, asset owners are aware of the challenges to the maximizing energy efficiency and also know that achieving energy efficiency will come at a cost.  If low-cost technologies would be developed and made available to help asset owners to achieve energy efficiency, it would urge the asset owner to want to own such technologies and achieve energy efficiency.

MG: Yes; energy spending within the building is typically a building owner's greatest expense. Not only is there financial pressure to adopt energy efficiency, but social pressure, as well, to increase sustainability and do what's right for the environment. Our customers are aware that energy is not limitless, and they want to be good stewards of our limited resources for both today and tomorrow. Their challenge, though, is typically around what to do to achieve energy savings. Having the domain expertise to look comprehensively at a building and then define all of the potential opportunities is difficult, even with the advent of IoT-enabled software. Helping a customer prioritize the opportunities and then go further, executing on the opportunities, is where we find our customers want the most help.

GC: How early are we in the IoT journey, in particular? How well aware are asset owners of connected solutions that can help address energy efficiency in their buildings?

DB: I think the IoT journey is still early in its development, with a need to move from a plethora of technical approaches and one-offs to a more coordinated, standards-based movement. Ones that can better leverage open source, cloud, and container technologies for speed. It is not about single vendor lock-in, but more about how we aggregate the best capabilities to deliver the desired outcomes.

MO: I agree with Dave; ubiquitous sensors and networks are the foundation, but IoT applications that move from cloud to the edge and back using container technology are necessary to address the IoT challenges in energy and other industrial markets. 

YN: I agree. The industry is still at the planning stage of the IoT. The market is developing new technologies (e.g. low-cost sensors) and embedding them into building equipment, devices and appliances that will support and integrate with the IoT. Almost all homes and buildings have some sort of mini network (e.g. Wi-Fi), which equipment and devices can leverage and ride on to transmit energy data over and to an IoT cloud.  

MG:  I would say that commercial and industrial building automation systems (BAS) have had IoT connectivity since the early 2000s, so that aspect is further matured than most other building systems. Where I agree with my colleagues is that taking the vast data available from a building, determining the proper use of the systems and generating optimized sequences is still in an early area of development.  Whether it's defining energy savings or further optimizing the many hydronic and airside systems in a building, this discipline is still growing.


GC: In a world of limited resources and a need to prioritize, among the array of connected solutions available to building managers, what do you see as the most significant for improving the energy efficiency of assets? What has or can be the material impact of installing these smart/connected systems (such as smart meters, connected HVAC, etc.) into commercial and residential buildings?

DB: The new generation of smart/connected systems coupled with more contextual data points, such as those found in new LED lighting platforms, is driving strong ROI in one to two-year timeframes.

YN: PG&E smart meters may be leveraged by our customers and can be used as a gateway to facilitate and promote energy efficiency.  The most significant improvement in building energy efficiency may be seen with new lighting technologies, new energy control sensors, devices and applications. This will take time, as affordability is still key in growing the use of new devices.

The most significant improvement in building energy efficiency may be seen with new lighting technologies, new energy control sensors, devices and apps
— D. Young Nguyen, PG&E

MG: I definitely agree with Dave that LED lighting is a "game changer" in energy efficiency for buildings. Not only is there an energy savings aspect, but the longer life of a LED fixture means that there are real operational savings from avoiding re-lamping and maintenance. Further, I believe that the pervasive increase in residential IoT-connected devices is helping the homeowner to have more choices in how they manage energy usage.


GC: For GE and Trane – what has been the key challenge you’ve seen in actually creating a connected solution? What can the ecosystem do better in order to better facilitate the IoT solution development process? Also, how has connectivity changed the way you look at your business?

DB: A challenge we faced was the lack of a PaaS offering that could address the breadth and scale of issues across the built environment.  This is why we invested heavily to build an Industrial IoT platform with the help and guidance of the deep domain skills of our engineers, working side-by-side with our computer and data scientists. This platform, Predix, enables app developers across the greater ecosystem to leverage catalogs of analytics and participate in solving pressing issues. Participation and joint collaboration across the ecosystem will amplify our innovation and speed our time to market.

As far as connectivity is concerned, we need to abstract our connectivity above the physical mediums, approaches, and carriers to assure the right QoS and time series correlated data sets. Software-defined networking helps us to do that. At the same time, we are investing in the latest ZigBee mesh networking to move with scale and speed across the light fixture-enabled sensors at the right cost point. 

MG: Interoperability is—and will continue to be—a challenge in the coming decade. The speed to which manufacturers are running to have IoT-connected solutions is rapid and this means that everyone is finding and/or developing the technology and then quickly going to market. This offers later challenges in connecting various products and services. We all have a stake in data integrity and security, so continued attention in this area is expected. That is a real concern for customers today. 

Connectivity has allowed us to further our R&D with real customer data. Engineering can see actual performance of HVAC systems not just in a lab, but also at the customer’s site
— Matt Gates, Trane

Connectivity within our organization has allowed us to further our R&D with real customer site data.  Engineering can see the actual performance of HVAC systems not just in a lab, but then later at the customer's site. We also have implemented our own building energy management software (BEMS) in our facilities. Our parent organization, Ingersoll Rand, has pledged publicly to improve climate conditions at our own operations and with our products, and we are using our BEMS technology to reduce greenhouse gas emissions and improve efficiency of our operations.


GC: For PG&E in particular – what has been the material impact of installed, connected systems? How has this changed the way the utility operates? How has this helped your customers?

YN: PG&E helps its customers save energy. PG&E’s energy-efficiency program currently provides multiple options for customers to achieve their energy efficiency goals.  The current options may include incentives and rebates for efficient lighting, building insulation, air conditioner peak time curtailment, Home Area Network, etc.  With the IoT, PG&E will need to develop and/or leverage new technologies to be in a position to better support and serve the customers.  And this will change the way PG&E operates, collaborates with the industry in deployment of new technologies, and serves its customers.


GC: For Frost Data Capital – analytics are the 'secret sauce' that takes the value prop of connected systems to the next level. Where do you see early applications, and what has been the impact?

MO: How about this logic? With more sensors we can measure and generate more data. And more data allows us to derive more actionable insights. And with more insights, we can make buildings and even cities more energy efficient. So, we simply need to put the technology in place and the world is a better, greener place, right?

Not really. This train of thought is flawed and contains a common misconception about IoT. And it’s a big one. Throwing more sensors at a problem and pumping data into an “IoT” cloud platform is like handing your phone to your dog and expecting it to call the pet groomer to make an appointment for a lion cut. It’s not going to happen.

Throwing more sensors at a problem and pumping data into an ‘IoT’ cloud platform is like handing your phone to your dog and expecting it to call the pet groomer
— Matt Oberdorfer, CTO, Frost Data Capital

Why not? The dog is given a phone with all the data and even the possibility to act on it, but it will most likely just chew on it. Because it doesn’t ‘know’ what to do with the phone “thing.” 

In every IoT project, the “intelligence” that is needed to process the data does not just magically appear. It has to be created before it works. And typically that involves a data scientist that truly understands the domain knowledge and translates it into a model. That model has to be fed, trained and tested before an app developer takes over and builds a IoT analytics application prototype… which then has to be field-tested. This process can take months or longer. And that’s true for almost every single IoT project, because almost all are unique in nature.

But, what about machine learning? Isn’t that the Holy Grail for analytics application that will solve it all? Well, no. It’s just a tool. A carpenter uses a hammer (and other tools) to build a house; machine learning algorithms are the tools that data scientists use. The tools don’t create solutions; people do. 

GE and a few other companies understand this. For example, GE’s Predix platform is not just an IoT platform, but it’s an ecosystem for data scientists and developers, as well as a marketplace for analytics and app-serving partners and customers. 

But if you really believe the numbers, we are facing an even bigger challenge: If in the next five years the number of sensors and devices grows from five to 20 billion, will the numbers of data scientists do the same thing? That’s where the logic breaks down, because even today there are many more data science problems than data scientists.

So, what can we do? We have to find ways to give “super powers” to existing data scientists or automate their work altogether. That’s where we see the IoT challenge. That’s why we invested in companies like UBIX and ThinkIQ that are working on future technologies like Automated Data Science.


For much more from this fascinating conference, visit the IoT World YouTube page

Google+ Google+