ASHRAE Journal:
ASHRAE Journal presents.
John Falcioni:
Hi everyone, and welcome back. This is ASHRAE Journal Podcast, episode number 17. I'm today's host, John Falcioni, the editor of ASHRAE Journal.
It's no surprise that in the technology rich world in which we live in, we are also seeing a surge in artificial intelligence within building systems and devices. This is the topic of today's podcast. This use of tools and technologies which are part of the Internet of Things ecosystem is impacting the way buildings and campuses operate and affecting entire cities and residents. If it's done right, AI technology has the potential to significantly improve the experience of building occupants and also to increase operational efficiency and optimized space and asset use. So what's happening in essence is that some buildings have become more than simple walls, roofs and masonry. AI is empowering building systems to autonomously integrate buildings data from IoT devices, and also capturing data on how occupants behave. This data which turns into machine learning can optimize building performance and improve environmental efficiency. The information captured from digital devices generates the intelligent insights about the operations of the buildings and the use and the condition of everything from a building's infrastructure to its physical environment, its climate, and its water and energy usage.
As I said before, it also delivers information about the experience and satisfaction of the people in the building. Sounds amazing, right? But are we really ready for the full integration of this technology?
With us today to talk about the opportunities and challenges that come with artificial intelligence and machine learning, in building design and operations, and also about the issues facing the industry in accepting these new technologies and tools and also preparing the next generation of AI savvy engineers are Dr. Zoltan Nagy, the director of the Intelligent Environments Laboratory at the University of Texas at Austin, and Dr. Rania Labib, who established the Artificial Intelligence and High-Performance Buildings Lab at Prairie View A&M University. To read their full and impressive bios, I invite you to visit the ASHRAE Journal Podcast page on ashrae.org.
I want to thank you both very much for joining us today. We've got a lot to cover. So let's begin by trying to contextualize what artificial intelligence means in the building industry. In other words, how do we define AI in the design and operations of buildings?
Zoltan Nagy:
This is a very big question to start off, John, but it's a good one. First, I think I would take a step back and say why we are thinking about this at all. I think, like you said, buildings have a big impact on energy demand, but also emissions. And so anything that gets us towards building better buildings, operating buildings better, reducing our carbon emissions from the built environment is a welcome technology. And so this is a next step. And so I think AI or advanced techniques, advanced computational techniques, have many places in many different areas. We'll cover some of these throughout this podcast. You can start by how can techniques support the design process, the engineers, the architects, how can we be supported during the operational phase? And then also how can we think about deconstruction, I think throughout the whole life cycle of the buildings to make sure that we really, really bring out everything that we can.
John Falcioni:
Excellent. Rania, would you like to add something to that?
Rania Labib:
Artificial intelligence and buildings is, we have numerous opportunities with the tons of data that we have within the built environment. So there are numerous opportunities for AI to improve the building performance in the comfort of the building as well. So many of the buildings are not comfortable in terms of thermal comfort or visual comfort, but these can be improved with AI technologies that are a designed to be custom for the human comfort that are inhabiting the buildings. So we have numerous opportunities right now with the available data through sensors, and we'll talk about this later also on the podcast. So there is a good chance that we can reduce the energy and consumptions and buildings, and we all know that based or according to the US Energy Information Administration, building energy consumption accounts for 30% of the global energy consumption. So this is a good time to utilize all these AI techniques to reduce the energy consumption.
John Falcioni:
That sounds great. So what is the promise of using this technology? I know I mentioned a few things, but why don't you guys tell me what the promise of this technology is?
Zoltan Nagy:
Yeah, I think a lot of the vision that drives AI in buildings is a sense of autonomy, a sense of making sense from the lots of data that's being generated. Over the past years buildings, especially commercial buildings, they are basically data generators. We have lots and lots of BMS data from many, many points. And so it's almost impossible for a person to sift through that.
And so the problem is really in that area especially, is to say, "Well, can we sort of extract knowledge, see patterns, understand the building better, and then improve the operations?" And one specific example, which is also one that's probably the most advanced that is being used is automated fault detection, where we sort of try to figure out how buildings should behave and then we monitor it against that. And then we can raise flags when the building is not behaving. Whether that needs AI or not is a different question. The most advanced techniques do it. You can do it simpler, but the more and more data we generate, the more complex our systems become, the more sophisticated the monitoring techniques and the analysis techniques also will become. So that's one big one.
John Falcioni:
Zoltan, you mentioned the sort life cycle of an operation of a building. Why don't we begin digging into that? Why don't you talk to us about the three phases that you discussed earlier?
Zoltan Nagy:
Yeah, I think, like I said before, so we can talk about design, we can talk about operation, and then generally the whole life cycle of a building as we talk about AI techniques, a lot of buzzwords around these days on digital twins, which tends to try to link all of these together into a digital model of the building, which encompasses the BIM part, the architectural drawing part, the operational phase. And so it becomes a living twin or a ghost of the actual building that we can use then with computational powers to improve in the operation.
I'll let Rania speak to the design phase, but I wanted to give another interesting, I think one of the big opportunities for advanced techniques in buildings operation is the idea of customization or personification. We operate currently buildings to set rigorous standard that have very little to do with how the occupants actually use the buildings or how they are actually feel comfortable in the buildings. And those are set to whatever standards we have, but they are not really the way the buildings are being used. And so there is a lot of mismatch between how building is used and how it's designed to be used.
And so using data that's being generated during the operational phase, like when do people come to office, how do they use their thermostats? How do they open their windows? How do they use their lights? It's different in every part of the building, in every building that you go into and using that knowledge to improve the operation, I think that's a big opportunity. And you need some advanced techniques because it's so much data that you can't sit down and look through it. That's basically the premise of that, the big promise, premise and promise.
John Falcioni:
So Rania, why don't you talk to us a little bit about the design process and your feelings on the impact of newer technologies.
Rania Labib:
Okay. So in early design phase of a building, design decisions have the most impact on the final building energy consumption and the final building cost as well. So it's very crucial to design in the very early design phase. To support energy efficient design, architects have started to use or utilize a set of performance simulation tools like the lighting simulations, thermal comfort, energy modeling, all these kind of simulations. These simulation tools are often integrated in parametric environments where people started to produce all these parametric buildings, hundreds of configurations in the early design phase, all these things, all these cool things. So they control the parameters to produce various configurations, but to choose the most optimized configuration, it's very challenging because you have to simulate the performance of all these configurations. So it turns out these are time consuming tasks, computationally expensive tasks as well. Not a lot of architecture firm have the right equipment or hardware to complete the simulation tasks.
So here comes the power of AI really. So if we can integrate AI tools and the building performance simulation tasks, it can be very powerful, because it can reduce the lead time to almost instant. So instead of waiting weeks and weeks on a decision or on simulating the performance of one building configuration, you can have the instant prediction of the performance, and you can choose the ultimate or the energy efficient building design configuration. So let's take for example, daylighting performance simulation. Daylighting, by the way, according to the US Department of Energy, artificial lighting use in buildings can account for 17% of the building energy consumption. So if we can reduce the dependence on artificial lighting, we can save up to 17% of the building energy consumption, and that equates to 186 tons of carbon emission as well. So if we can design proper daylighting system, we can increase the energy efficiency and we can decrease environmental pollution as well.
So daylighting performance simulations can take two hours for one building configuration. So if we are using these parametric, cool parametric tools and producing 100 configurations, for example, it can take weeks or months to submit the performance of daylighting of one building design. So it becomes a very time consuming that architects start abandoning all these tasks due to strict project deadlines. So AI tools, and I've published couple of papers on this AI tool that I designed for daylight and performance. AI can be integrated using all these simulation data. We can feed them to a machine learning algorithm that can predict for us instantly the performance of any building configuration that we want to use. There are still challenges. They’re not fully ready yet because most of AI tools right now are designed for traditional design, like a box building that looks like a box, where it looks like traditional walls, but not many buildings are designed this way. But there are potential with the advanced technologies that pop up every day on AI and artificial intelligence, we can improve these tools further and further.
John Falcioni:
Zoltan, I know that you've talked a lot about energy efficiency. Would you like to add a little bit to what Rania is saying?
Zoltan Nagy:
Well, I think the promise of, again, AI is to make sure that we can cover our bases, really. We can explore so many options that normally we cannot. One thing I will add to it is that, and you noticed when I used in my, when I talked about is I said, "Advanced techniques" and I say things like, I avoid using AI or machine learning because it triggers a fight little response from many, many people. And so I will just say this, it's not necessarily any sort of AI, it's also anything that has to do with computation.
The other things we need to make sure is that just we build better. It's not just to use technology for the sake of technology, but to really make buildings better, even from the design onset. And what it can help us is to cover a lot of variations that we just can't handle by hand or in our brains. And so we can use the power of computation to explore that. And so the way you want to think about it, I think when it comes to design, you don't compare an architect or an engineer with the AI, you compare an architect against an architect with AI support. So it's more like decision support system when it comes into the design process rather than a replacement, because that's the way things are moving and have been done in other industries as well. And so the building industry is lagging and slowly catching up.
John Falcioni:
Well, yeah. I was actually just going to ask you about that based on what you just said. How do you feel that the building industry compares with other industries in the adoption of technology, forget AI, but just all types of technology?
Zoltan Nagy:
Yeah, I think, like I said, it's lagging, which is the nature of things, the industry is slow. It's slow to adapt innovation, that's how it's always been. It's nothing new. So it's just another cycle of these. But just, it didn't use to use CAD, then it used CAD, then it didn't use BIM, now it's starting to use BIM, so it's the rule of things.
Other industries can innovate faster because they have shorter innovation cycles. So if you build an airplane, you build many, many, many, and they are all the same. And so you can improve them into a computational model makes sense because you can improve in silico in the computer model. If you build cars, the car cycles are a lot shorter, so you can invest in those tools. Whereas in building industry, it's hard to see the investment in advanced technologies to justify. I think that's been a challenge for a long time. I tend to think I'm not that old, but I'm pretty sure this has been an ongoing challenge in the building industry for a long, long time.
John Falcioni:
You mentioned in one of our previous conversations, Zoltan, that every building is a prototype. Can you tell our readers a little bit about that perspective?
Zoltan Nagy:
Yeah, I think this follows very neatly from that. When you build a building, you think about it as a unique thing. It's not like a consumer product in many ways, right? In every single building that you build, even if it's the same plan, it's almost, even if it's the same construction crew, you will end up with two different things. And at the end of the day, it will be inhabited with two different families or occupants or tenants. And so at the end of the day, it's two different things in that sense. And so that's again, the power of computation is to— in case you are going into operation to make sure that everything can work on different types of buildings and all these things that are similar, but this different, trying to understand how things can work is again the part where AI can help.
John Falcioni:
Rania, you talked a lot about workforce when we spoke before, and I know that that's one of your fortes. The implementation of these technologies to the buildings industry is the lack of that integration part of a workforce development issue? Is it because students don't come out of school really understanding how to use these tools or adopting these tools?
Rania Labib:
Yes, definitely, workforce development is a part of the issue. Because again, AI technologies to truly create an effective and transformative AI technologies, this requires multidisciplinary efforts. So multidisciplinary efforts incorporating not just engineers, computer scientists, but also architects. The problem is in architecture, and architecture school, and sometimes even architecture engineering, and maybe Dr. Nagy can elaborate on this. I'm a professor in architecture, but architecture schools and architecture curriculums have been very traditional, very slow to adapting to any technologies, let alone artificial intelligence. So that's a very big problem, even though there is a huge demand right now for architects with knowledge and embedded technologies, embedded sensors, IoT, AI, all these things. But again, architecture schools have been very slow and traditional.
In the I, myself as an educator, as a professor, I see students that are reluctant to learning these technologies. They think, "Okay, we are here to design. We are here to design a very beautiful model of a building in a BIM platform. Why do I need to know about AI technologies or even just technologies in general?" Any technology. They think of themselves as creative designers, not really AI and technologies are just something far reach from them and they don't want to really approach it. So they think it's hard and it's difficult. And they don't realize the power of these things, the power of these tools that they can use to improve their designs. They don't really realize that.
So this semester, for example, I took an initiative, a new initiatives, I'm not sure, it's still experimental, but I teach a lecture class along with the studio. So the lecture class, in place of formal lectures, I started teaching about AI stuff. So in the first week we watched videos about AI technologies. The second week we watched technologies about IoT and AI in the building industry. And they started being interested. And then I walked with them through the creation of a small AI tool where they designed it under my supervision, of course, they are not coding people. But under my supervision, I taught them how to design a small tool that can help them make decisions quickly in terms of daylight and performance for a shoebox model, like a small room.
The experiment has been successful so far that four of my students are advancing to enter a competition that is supported by the US Department of Energy. The competition is called the JUMP into STEM. So they are going to enter this tool that they designed with me into the competition. We are going to fine tune it a little bit and enter it into the competition. So if we can spark the interest of these architects right now, that's wonderful because later in the future, we going to get overwhelmed by these technologies and we are not going to be able to handle it. So they need to be prepared for it. They need to be prepared right now, because it's happening in a few years. A lot of the things around us will be embedded with AI technologies and it's going to be hard and overwhelming for them.
ASHRAE Journal:
Seeking the job in the engineering field or searching for the most qualified engineers? The ASHRAE career center connects opportunities and candidates. Search for jobs, manage your resume and create alerts for when jobs are posted. Employers, post new jobs, review resumes, and manage recruiting. Go to jobs.ashrae.org.
John Falcioni:
So there is plenty of data available right within the built environment, but limited access to the data to use the data. Again, I'm thinking that this is part of that whole workforce development issue. Zoltan, have thoughts about this?
Zoltan Nagy:
I think data availability is one of the biggest challenges that we have in AI. And if we think about building operation and developing methods to improve building operations in a way that scale across many, many buildings, one thing that we need is data that is from real buildings. And this has been a problem or a challenge I should say, because what you are asking is building operators to share their data. And it has two layers. One layer is how, and that goes towards what we discussed on, '"Yeah, we need improvement on workforce development to understand data, what is collected." It also needs improvement on the technology side to some extent, to actually store the data. A lot of the data is generated but never stored, which is basically throwing away a little bit of money there.
And then at the third, it also needs some, I think, rules. How do we share that data in a way that's comfortable for everyone? You can learn a lot from a building by looking at its energy data, especially if it becomes high frequency, five minute, 10 minute, 15 minute, and many, many points, data from different rooms, different areas of the building. So depending on buildings, it can become a safety issue, it can become a privacy issue and all of those things. So those are more regulatory that needs to be handled so that it doesn't become too exposed or at least make the building owners comfortable with. There are a few positive developments over the past few years within a lot of different professional societies. DOE has led a big project on collecting data and making it available. IEEE, ACM, all generate task forces for data because they know it's important to develop the next generation of tools as we head into more energy flexible buildings and all those cool things.
John Falcioni:
Yeah, that talk about data sharing I'm sure scares some of our listeners. So how can we make the listeners feel a little bit better about these problems? Is there a way to—
Zoltan Nagy:
There is in fact because you only need to think about, go back to your first iPhone that you own, maybe, I don't remember, 20 years ago, 15 years ago? And if you took a picture, you took a picture. But now if you open your iPhone and you look at the screen while you are trying to take a picture, it will show you what the heads are, it will track the people, it will improve. So the technology has improved quite a bit. One of the reasons it has improved quite a bit is, well, first computationally, it's easier to do a lot of these cool things than it was 20 years ago.
But the other reason is because at least in that field of image detection, there has been a lot of progress through shared data sets. The ImageNet that came out a long, long time ago was used as a benchmark data set to test the algorithms in terms of their performance, in terms of their accuracy, in terms of their computational power. And that sharing of the same dataset has led advances or has led to advances because people could experiment on the same dataset. So what we are trying to see in the built environment or what we need in the built environment is a similar approach where we can test advances and algorithms on the same data sets. And so again, this has been this course that we need dataset. I have been on that board for a long time, so it's good to see that we are seeing these datasets coming out over the past couple of years.
In fact, ASHRAE had a competition just a few years ago on the big Energy Shootout, the third edition led by Clayton Miller. And we contributed a little bit to that as well. And the idea was to generate a big data set where a building that can be shared so that people can compete on energy forecasting, for example, in this particular case. And that dataset is now available so you can compare. And it also has led to understanding a lot of things by reusing the dataset and working on it and tackling different problems. We now know a lot more than we knew before on things that work and things that don't work. So these type of initiatives are needed. That particular dataset is very specific, but it has shown us that it can work. We need more of those.
John Falcioni:
Can you tell us about your intelligent environments lab and city learn initiative? I know that there is a competition involved there as well, correct?
Zoltan Nagy:
Yes, correct. So my group, Intelligent Environments Lab, we focus a lot on this operational efficiencies. How do we operate buildings efficiently? One of the topics that became clear to me early on long time ago is that building will be used more to provide flexibility in the energy space to the grid. Especially when you think about a lot of, let's say single family homes that run their thermostats at the same time, they charge the batteries at the same time, they probably will plug in their EVs at the same time. So while individually they will become very efficient in using solar batteries, vehicles and whatnot, collectively, they will just move the demand to another time and may not actually save us to reduce the peak grid demand.
So what we started thinking about is how could we collectively or coordinate demand from buildings. In this case, mostly single family because that's how you aggregate them up. And so we started to build this computational environment that lets us do that and study. And so that individual buildings don't necessarily do the same thing, but they coordinate together to reduce the overall demand from the grid. And so over the last few years, while we made progress, the other buzzwords that came up were grid interactive buildings, connected communities, all those buzzwords are integrated. So there is a lot of momentum in this field.
And so again, we did three competitions. We did two, 2020, 2021. And this year where the idea is to, here's 20 buildings this year in California, we only give you access to five. You have to develop your controller and then we'll deploy it on the other buildings and then see how they perform. And we won't judge you on how they individually perform, but we judge you on how they collectively perform to reduce the emissions from the grid, to reduce the cost to the owners. And all they do at this point, for this particular example is charge and discharge the batteries while they are using solar panels. And we already see that the simple solutions don't work. So the solutions that we typically have when you install these in your homes is to charge in the morning, discharge in the peak afternoon. So from 2:00 to 8:00, or from 3:00 to 7:00 or something like that.
But if all the buildings do that, then you end up with very high loads at the end of the day when everybody starts to ramp up. And so that's what we are trying to avoid. And so spacing those out, those charging and discharging times really helps. Even in this very simple example, this is only 20 homes with the battery and solar panels. It's not super advanced. But what if happens if you add to that homes that also have vehicles, then that battery that's in the vehicle shows up or is not there at times. So it's a much harder problem. What happens if you have not 20 buildings, but a hundred buildings and some of them have PVs, some have batteries, some have none. And so you have to try to maybe shift the thermostats a little bit, but in a way that doesn't bother the occupants. So that's always the baseline of these. There is no impact on occupants. And so that's what we are exploring in these competitions because it's a lot easier to get a lot of solutions very quickly when you let people compete against each other.
John Falcioni:
And Rania, you run the high performance buildings lab. Can you tell us a little bit about that?
Rania Labib:
Sure. So I established this lab about a year ago, so it's fairly new. The lab is in the school of architecture in Prairie View A&M University. It's part of the Texas A&M University system. So it's sister of the Texas A&M system. The major drive of establishing this lab was to improve AI education and architecture school that leads to the design of high performance buildings anyway. The lab, by the way, has the state of the art machine learning workstation that anyone is welcome to use under my supervision since it's a little bit of expensive piece of equipment. And we have access of course, to the high performance computing facility through Texas A&M, it's called or known as supercomputer. In the lab, I introduced the students to a lot of data collection concepts utilizing many equipment. So some of them is the LIDAR equipment, drone-based equipment like thermal imagery and other meters that can measure moisture, humidity, air quality, light level, and et cetera.
Students are also introduced to AR and VR technologies, which are not necessarily related to artificial intelligence at all, but this can be something that can prepare the students to become digital natives and can help them in the future utilize any AI technologies. So one of the projects they have completed as a student project was to fabricate a complex structure using steamed wood. But to bend the wood to follow the structure curves, we had to use AR glasses where they can see the full model that they created in the modeling environment. They can see it in a real life as real life-size structure, and they can start bending the wood to follow the curvature of this structure. So students had a blast on this, really became very interested. Some of the students who operate my drone as well, they became so interested that they bought their own professional drone, not research grade drone, but professional drone, not a toy drone. And they became very interested in collecting data.
Of course, we have some challenges. And the research, I'll talk about challenges later, but let's talk about research effort in the lab. So I'm working on two research projects right now. One of them is with collaboration with Texas A&M University, where we do machine learning tools. We design machine learning tools to predict thermal conference within existing and in new cities or dense urban areas. This is still in the early phase. The other project where I'm having challenges is we are trying to collect drone imagery paired with infrared imagery as well. But we are facing with the FAA regulations or the Federal Aviation Administration regulations, where we cannot scan some residential buildings due to privacy issues of course, until we have permission from these houses. But basically we collect images and pair them with other information to predict and feed them to a machine learning model to predict the urban heat island effect. So this is going a little bit slow because of all these challenges. Maybe we'll try to reproduce other information besides images, because images has been very tough to obtain. We can obtain them, but due to the strict rules, of course.
The lab, again is still new so in the future I plan to utilize a lot more of sensing technology as an IoT sensors. Specifically interested in utilizing three dimensional cameras where we can collect the information of a person in a space, so the coordinates and where that person is in the space, and also recognize their facial aspects so we know what person, who that is and where in the space, that will help us monitor the occupancy patterns within buildings. And this gives us insight and into thermal comfort of the building or the visual comfort of the building and other aspects of the building. I'm hoping I can get permission for these cameras within my school, at least as a test ground. So still also in the early phase. So a lot of challenges I'm facing, but I'm hoping this is a new concept in architecture school and I wish it is a successful lab that can be reproduced or mimicked in other schools, hopefully.
John Falcioni:
Let me circle back a little bit to the energy issue and energy savings. I'm not sure if we fully talked about the potential energy savings that this technology can generate. You talked about 30% of the total energy coming from buildings, Rania, is that message, do you think loud enough for those who worry about these things to understand the importance of this technology?
Rania Labib:
Yes, definitely. It's a big message. And also with the urbanization that we are facing, the growth of urbanization, we have to act fast and we have to take advantage of all these available data sets, AI technologies to improve our future building. So we can collect a lot of data on our behavior as occupants in the buildings, how we use the buildings, how these buildings act to these occupancy patterns. And we can design better buildings in the future.
John Falcioni:
Zoltan, anything else to add on the energy comments?
Zoltan Nagy:
I think one thing that we ought to understand is that we cannot operate buildings the way we used to. And that buildings will become an intricate part of the power grid system. And that is because we try to move away from fossil-fuel fueled power grid, I mean we should. And so what it means is that we need more renewables in the grid. And whether that's solar, whether that's wind, nuclear, still a little bit of fossil fuels, the demand that the buildings have will have to adapt a little bit to when power is available. And that can be done in very different ways. One is active and passive storage, so batteries, thermal storage, electric batteries, operating the buildings with pre-cooling, pre-heating, things like that. But essentially it's different from the way we used to which was, "Do whatever you want. If we need more power, we build more power plants and we burn more oil and gas."
So this transition of aligning demand with supply at the same time that we are electrifying the end use or trying to electrify end use and adding new loads such as electric vehicles is a big challenge. And coordinating that I think is again, in operation where AI techniques will help us or at least advance computational techniques or we'll need a lot of computational power to understand what's going on, and to make sure that nobody is getting blackouts because we have an abundance of energy. It's just not at the time when we need it. And so it's all about shifting it a little bit on a daily basis, a little bit on a seasonal basis. But these are the challenges that we are facing. And I think that we are facing, we don't really know where it's going, but we are on that path.
John Falcioni:
So even without AI, is there more we can do to make better buildings?
Rania Labib:
Yes, of course without AI we can even improve our buildings. But it starts again with the early design process, which is part of my research expertise. So it starts in a very early design process. So we don't design huge buildings, huge spaces. There are even over the capacity we need and we start slapping huge HVAC systems on them. That it's overpowered and it just consumes a lot of energy. So we need to design the most efficient spaces in the first place. Our engineers, when they take the architecture drawings, don't start designing oversized systems on them. So the early design phase is very important. And again, and of course it's my research expertise here that I design all these AI tools to improve the way that we design in the early design phase, so we can produce the spaces that are comfortable in really high performance spaces that is the energy we can consume.
Zoltan Nagy:
I think there is a lot of opportunity to improve the building stock without anything fancy. So I think that's one thing we should take away from here is that what we are talking here about AI is coming, but it doesn't mean that that's the only thing, and it's not a savior. The building stock, especially if you look at single family homes, they are built really, really badly. They are built to the absolute minimum that a builder can get away. And that's not necessarily good in a sense of energy efficiency. It's good for all of those who sell you power because you buy more from them, but it's not good for exactly energy efficiency. And so basically we have a big undertaking to improve the building stock by tightening up the buildings, by making sure we transition away from gas furnaces to heat pumps, which will be possible once you start tightening up the buildings. So these things are not one or the other. It's both things at the same time.
What is interesting is sometimes you don't feel comfortable in your home because you are draughty or it's a little bit cold on this part of the home. So what you do is you crank up the heating or you turn down the cooling. But at the end of the day, if you actually looked at and let some professional check it out and you tighten it up, you may be getting away with the last little energy in both cooling and heating, and you could transition away from that dirty gas furnace that you have.
So I think bottom line is there is a lot of opportunities that we can and absolutely should do when it comes to operating the buildings to lower the baseline. And at the same time, that will lead us to match better the demand and supply of renewables. So it's all one big bag of a lot of opportunities, and AI is just one of those tools that in some particular good situations or coordinating loads, improving the design will help us more, but we have to cover our baselines. This is not a get out of jail ticket for those who don't want to make the buildings better.
Rania Labib:
So, yeah. Yes. I forgot to mention that Dr. Nagy, he talked about replacing gas furnace with heat pumps. So that's part of the decarbonization efforts. So the initiatives that has been going for years now. So in the part of the decarbonization initiative, not just to deliver the most efficient design, but also decarbonize the building and electrify our building. So replace all the gas furnace with heat pumps and possibly also add electricity generation equipment such as PV panels or other equipment to generate electricity, so this way we can reduce the carbon emission and increase electricity generation hoping that we can reach the net-zero energy design.
John Falcioni:
Terrific. So on that note, we'll end this episode. I want to thank our guests, Dr. Zoltan Nagy, and Dr. Rania Labib, for being with us today. I also want to thank all of you for tuning in. From all of us at ASHRAE Journal, I'm John Falcioni. Join us the next time for another conversation.
ASHRAE Journal:
The ASHRAE journal podcast team is editor John Falcioni; managing editor, Kelly Barazza; producer and associate editor, Chadd Jones; assistant editor, Kaitlyn Baich; and associate editors, Tani Palefski and Rebecca Matyasovski. Copyright ASHRAE. The views expressed in this podcast are those of individuals only, and not of ASHRAE, its sponsors or advertisers. Please refer to ashrae.org/podcast for the full disclaimer.