0:00 I didn’t touch anything. Welcome to our presentation on Aurora. Aurora is AI grounded climate intelligence for early stage design. My name is Ilya Glassman and I’m joined today by Demola Michael who’s getting his mic set up, but I’m going to kick things off for us. So my role is the as the lead of Cannon Design’s building performance group and Demola is part of our design technology team.

0:33 So just to give you a little bit of background about Cannon Design, we are an architecture and engineering firm. We have about 1,800 employees and about 20 different offices across the US and the world. And we work on a variety of different building types, but our key specialties are health care and education. So we see sustainability as one of our core living centered design values. We were one of the first signatories to the AIA 2030 commitment, which means that we have a goal of creating a carbon-free building portfolio by the year 2030, which is not as far in the future as it once sounded.

1:16 But you can see some stats here from our upcoming sustainability report that actually so fresh off the presses came out today, that we have a big commitment to sustainability. We have over 379 projects that have been LEED certified. There’s more coming down the pipe. We we have certified buildings in their other standards like well. We’ve electrified our buildings. We have renewable energy systems incorporated into our buildings and we have some net zero energy buildings.

1:46 So our goal in doing all this is to create a very regenerative design practice. So while sustainability is the responsibility of every designer and every engineer in the firm, we have a dedicated sustainability team that acts as the subject matter experts to help drive the efforts, set standards, and promote innovation in the space. So, my role is to lead a subset of that team, the building performance group, that provides simulation validated design guidance to support the different projects that are happening across the firm.

2:50 So, the backbone of the service is really computational design using parametric analysis and design exploration to connected to building performance simulation to study energy, daylight, thermal comfort, visual comfort, water use, renewable energy solutions, and much, much more to identify and generate design solutions that perform very well under multiple design considerations. In the time that we’ve been at the conference, it’s been really great seeing all the work that folks are doing in other industries thinking along these lines in terms of speeding up simulation, in terms of validating results, and visualizing different design options.

3:35 In different fields like aerospace, automotive engineering. Ironically, we’re the architects in the room and we won’t be speaking about geometry today. So, to to talk about what we will be sharing with you, I just want to maybe give a little bit more background about the connection between the built environment and the environment in case it’s not very well understood. Nearly half of all energy use in the US is not from buildings.

4:01 Sorry, is from buildings. And the processes used to generate that energy creates carbon emissions in the atmosphere, which in turn leads to changes in our climate. We have all these new words for for weather that we never had before, including one we’re experiencing here, a heat dome. And that’s linked to changes in the climate, which create hotter temperatures, more extreme weather events, water scarcity in some places, yet flooding in others, and wildfires.

4:29 So, therefore, our goal is to really reduce energy demand and emissions from buildings cuz that’s a key component of helping to address climate change. So, the way to do that as designers is really to work with the climate where the building is going to be situated in. So, the first step is is passive design. We want to keep the building naturally comfortable, reducing the reliance on heating and cooling, and using natural light to offset lighting use.

4:57 The particular strategies that are needed to reduce all those things really do vary based on the climate that the building is in, whether it let’s say it’s a it’s a hot heat hot climate or cold climate. And that can be determined how much of the sun’s energy is going to be beneficial or harmful, how much shading you’re going to need. Outdoor temperatures and humidity will determine if strategies like natural natural ventilation or thermal mass will be beneficial.

5:27 So, we we do all these things to make the design of the building itself just need to feel more comfortable, need less heating and cooling. And once we’ve done that as much as possible, we meet those remaining heating and cooling and lighting loads as efficiently as possible with active systems. But even here the climate comes into effect because some of the system types, mechanical system types, or optimizations in the operation of those types of those systems, such as economizer controls or evaporative cooling, that really depends on the properties of the climate as well.

6:03 So, how can we help designers really understand how to use climate as far as their designs? Typically, designers could take one of two approaches when it comes to building design. One I call greening a building design, and another I call designing a green building. In the first, and sometimes unfortunately more common approach, the designers may let other considerations really drive the initial design, such as program, aesthetics, and cost.

6:30 All important considerations that should be incorporated, but it seems that only later they’re seeking to address the building performance. And the process then becomes one of you’re trying to triage the design to make the building building better, but this approach might be more costly because it might require adding more expensive things like better facade materials and glass, or more advanced mechanical systems, or renewable energy systems to compensate for the design decisions they’re making that might be working against climate responsiveness.

7:03 So, the second approach, designing a green building, is really starting off from time zero of the project with a really strong understanding of climate and what strategies you could use to to respond to it, informing things like the massing, orientation, and openings of the building. And and that’s complementary, and you look for the synergies between addressing things there in in terms of the program, aesthetics, and budget that we mentioned before.

7:31 So, that’s way more effective, and the costs are less since performance is embedded in the design. But to do this, the designers really need to be proactive at that time zero, and really understand the climate strategies before starting to put pen to paper. So, that’s where Aurora comes in. Hi everyone. My name is Demola, and I’m a computational designer. Right. So, as Andrew said, it’s very difficult, and it’s very time-consuming, and also costly to to try to green a building supposed to thinking about environmental analysis and sustainability at the inception of your building.

8:14 So, when does environmental analysis actually join the process? We believe environmental analysis should be part of you know, your pre-design stage. So, as you can see on the left-hand side, we’ve got a pre-design stage where you’re doing your site selection, you’re doing your visibility studies. That is exactly where environmental analysis issues start. Majority of architects and designers, they tend to like think about sustainability when you’re getting to the develop design stage, which can be a lot more costly and as well as Andrew was saying earlier, where people are trying to green a building supposed to making sure the decision that you’re making at the very start of your project has sustainability involved.

So, there are quite a lot of tools out there. We’re not trying to, you know, take away from great tools that already exist. But, the fundamental difference in also the problems with a lot of these tools is that they require you to actually have a model. When you’re doing your site selection, when you’re, you know, speaking to your client, you know, they just go site. They don’t know every geometry.

9:31 They don’t have an idea in mind. But, at the same time, they want to understand what the site condition is. You also want to understand what the environmental condition of your site is. But, you don’t want to have to go into, you know, go into Forma or go into Ladybug. With Ladybug, you have to go into Grasshopper or go into Rhino for you to do your analysis.

9:50 And that is where Aura comes in. Aura is a lot different because you don’t need no geometry. You just need to open the the web application and it gives you literally every single information you need regarding your site based on the site location that you provide. And there key sources where we’re getting our information from. There’s the EPW weather file that are coming from more than 3,000 global site locations, weather stations, where everything is, you know, we’re getting all those information and it’s pulling it through Aura.

10:30 So, Aura gets all of this information. The key one is the EPW file and it allows you to produce up to about 10 key analysis. We’re talking about your solar condition, we’re talking about your wind, your comfort level, your rain condition. These are key information that you need at the very start of your of your design. You don’t want to have to produce your geometry before you understand the the solar condition of your site.

11:04 Majority of the problem that happens is that a lot of architects, they start putting pen to paper, designing based on intuition as opposed to having understanding of a grounded understanding of the environmental condition of the site. And that is where Aura comes in, making sure by the time you provide the location of your site, you have all of the information needed to be able to put our produce a wall in front of a design.

11:35 And just, you know, comparing the way people work people work done in the past before we introduce our There’s that disjointment between the data that the architects work from and the data that the engineers work from as well. So, you can see the diagram on the left. This diagram is a classic problem. The conversation gap between, you know, the architects in the design the architects in the and the engineers.

12:09 The engineers talking about the architects majority of the architects they work in based on intuition, precedence, years of experience. A lot of the time they’re right or, you know, they’re close to being right, but sometimes as well they make a lot of mistakes. And we don’t want that kind of costly decision-making. We want to have a decision based on facts. Based on actual data as opposed to like intuition and that is where the problem starts.

12:42 So, having our, you know, just entering your site location it helps you to get all of this information without having to to think about precedence or using intuition to design. And as we know like in the AEC industry a lot of engineers are brought in towards the latest stage you were talking about developed design stage not at the concept design stage. So, that creates a a lot of problems because the engineers they don’t really understand the environmental condition of the site.

13:12 So, at the inception of your of your project you’re able to provide the same data that you’re providing to your client you’re able to provide the same data to to the engineers and also to to the architects. And if you look at the the below text rule that appear before you open anything. It’s going to save you a lot of time. It’s going to save you a lot of costly you know design decisions which earlier said in terms of like greening a building are supposed to making sure you thinking about environmental analysis and sustainability at the start of your projects.

13:57 And this is just it’s a video showing our works. So what you can see is guide it or you showing you like the wind analysis of provided. You’ve got a the main dashboard where you can do customize data arrangement because we know that not every data is going to be applicable. You don’t want to like you know over complicate you know your your decision making. So you can tailor your dashboard towards like the audience that you speaking to.

14:31 You know you speaking to your client or you speaking to the engineers so that I have seen that we’ve got a customizable course. So by the time you provide your location, there is a ready-to-use presentation or PDF document that consists of every single information we talking about your wind, your solar, your comfort level. It has literally every single thing that you are to present to your client and also to discuss amongst your your team.

15:00 And like I said, there’s the you know the key thing about the key feature of Aura is by providing just your location, every single thing gets populated. We’re talking about your actually design conditions that that that the engineers use. We’re We’re about sky condition wind. These are very critical information that knowing this at the start of your project goes a long way as opposed to like later down the line.

15:28 So, this just gives you an overview of what Aura can do in terms of other stuff just by providing the location that retrieves all of the data. So, we understand that there are some genius labs or like, you know, interns that you you give task to to produce some information or to do some research. But, not every single person is like Elliot the explorer that understands environmental analysis.

15:59 So, we need to make sure people understand what they’re looking at and that is where the analysis assistant comes in. Even though we’re making it so simple, we’re making it so easy for people to understand the condition of the site, we still need to make sure we’re explaining to people what they’re looking at. Because it’s one thing for you to have a chart and, you know, okay, what am I looking at?

16:25 So, and it’s another thing for you to have an assistant that helps you to understand exactly what you’re looking at and that is what the analysis assistant does. It also functions as the the main runner of the of the application. So, you can be you can basically run the entire application from the analysis assistant telling you to change your site location. If you go to different site, you can paste your site location there and it’s going to change the address for you.

16:52 It’s going to populate the necessary information you need in terms of the data that come for rain and the lights. It does that for you. It basically runs the entire application. But, it it’s very difficult for you to for you to ground an LLM because we’re using we’re using Gemini for our analysis assistant. So, we’ve tried as much as possible to make sure we’re not giving it any room to hallucinate.

17:25 So, we’re making sure the data that we’re gathering, like I mentioned earlier, you’ve got about five different key sources that we’re gathering to to provide to our We’re making sure we’re packaging everything very well in a context builder to make sure by the time it’s handed over to Gemini, it’s not hallucinating, it’s not making up data based on its training. It’s basically, you know, the information that we’ve packaged that we’re attaching to the to to Gemini.

18:00 That’s basically where it’s getting its information from. So, it’s not having any room for it to produce like irrelevant information. So, we’re making sure everything is fact-based. And at the same time, you know, like I mentioned, it can be very difficult to control, you know, the explanation of the results of the LLM. But, one key thing to know is that the charts being produced, those are accurate.

18:29 So, either way, you have the ability to be able to, you know, have a look at the charts as well as the the results given back to you by by Gemini. There’s so many improvements that we need to make. There’s so many features that we also need to to add. There’s some some that we’ve already started that we’ve got in the pipeline, and others are I will see you kind of like, you know, do research, and then Ellis is going to talk about the the future plans.

19:06 Yeah, so we continue to evolve over as we integrate additional features. This one’s not so much a future feature, it’s like now it it in the past few days we’ve we’ve built it in. While we while we were doing this presentation, the So, what this does is it based on again just one bit of information you put in the project location, it will collect the the utility costs and from and that includes water, gas, electricity from the local utility and now bring that as well as grid carbon emissions data for the location.

19:44 So, this data is actually really important, too, and it usually comes from a different source of information than the climate data. So, it’s a great that it’s aggregated in one place because it helps us understand the economic feasibility of different design features and how we can maximize environmental impact. Because then we could compare it to US average and understand is this place more carbon intensive and that informs whether like what we use for electrification strategies and decarbonization strategies, what are are the relative energy costs compared to others, so we could contextualize the results and help us better determine that strategy.

20:29 And there’s also another source of data that will pull in for hourly grid emissions, right? Cuz if you use 1 kWh of electricity, it’s not the same impact environmental impact at different times of the day or at different times of the year. So, we’re pulling in also the the hourly grid emissions from various sources and looking at future grid projections as well so that we can investigate and recommend strategies related to shifting time of use for the energy and thinking about what the grid is going to be doing in the future.

21:06 So, just to sum up, the understanding of the climate and site considerations really critical for informing an environmentally responsive approach to building design. So, what we’ve done is use our run develop it as a web platform to automate the collection of existing data, which typically would take a while from a wide variety of sources, which are constantly updated. And then and we synthesize that information to a series of of dashboards.

The factors captured include current climate, future climate, and aspects of those related to temperature, solar radiation, wind, humidity, rainfall, and many more. And the graphs can be already sent and exported as a PDF for deliverables. So, once you get that instantaneously as opposed to something that would take some days to do, which helps us move at the speed of the design and align everybody early on in the process.

22:02 What is that we’re trying to do is get so, our explanation but the AI AI is to help project teams better understand the implications of data even before they start putting pen to paper. Once they do, we have other tools that can help with the the design and evaluation, like one I briefly showed earlier. But, this is really getting in at at time zero. And the platform will continue to add new features to provide a better understanding of key priorities in relationship to the specific considerations that we just talked about in terms of utility cost and hourly greenhouse gas emissions. So, thank you. That’s it.