Scaling Sustainability with Autodesk's New Sustainability Data API

LUKA KRALJ: Hello, everyone, and thank you for coming to this talk. My name is Luka and I'm a sustainability platform specialist. I work with the sustainability solutions team under the wider Autodesk impact organization. My background is in computer science, so my main focus is on prototyping and delivering new sustainability solutions with the focus on our platform.

Today I'll first talk about what sustainability is and what the challenges are, and then you will learn how we are thinking of leveraging our Autodesk Platform Services, or APS for short, to make our and your sustainability efforts more efficient. Before I begin, I just want to flash this slide on the screen, which I'm sure you're well familiar with by now. So the work in this presentation is more research oriented and in the proof of concept stage, so what I'll be talking about may change in the future as we learn more about these topics and also learn more from your feedback.

Therefore, you should not make any purchasing decision based on this talk. In today's talk, we will first look briefly at what sustainability even is and why we should care. I will primarily focus on sustainability in AEC just to demonstrate the challenges around the sustainability data. However, it's important to understand that manufacturing is facing some of the similar challenges, as well as some of the ones that are unique to that industry. If you're interested more about the manufacturing side of things, I'm happy to chat after the talk.

Next, we'll dive a little deeper into the complexities of the sustainability data sets and what technical challenges they pose. We'll then look at what Autodesk is doing to tackle some of these challenges, and we'll also see how you can help contribute to our sustainability platform. The goal of my presentation today is to encourage the conversations around these challenges and working together to achieve a more sustainable future.

Without further ado, let's take a look at why sustainability, why should we care, and is sustainability just a buzzword. Unfortunately, it's not just a buzzword. Climate change is a global emergency that goes beyond national borders. And while you may have heard about these numbers before, it's important to understand why this is so important for Autodesk and the industries we serve.

Because we realize the importance and the urgency of this matter, Autodesk itself has been one of the most sustainable companies for several years now. However, as ironically as it may sound, we also know that we serve some of the very heavy polluting industries. Architecture and construction, or AC for short, and design and manufacturing, D&N, together generate around 60% of annual global CO2 emissions.

Autodesk has a significant market share in these industries. And because of that, realized that if we create solutions that help you create more sustainable buildings and optimize your manufacturing processes, then together we can create less emissions in the futures in the future. Now let's look concretely at a few examples of how we calculate some of the sustainability parameters and metrics in AEC space, just to demonstrate some of the challenges when it comes to modeling the buildings and getting the data that we need to create a sustainability analysis.

In this talk, we'll focus on the carbon. And carbon is quite a broad word, so I wanted to break it down really briefly what I mean when I say carbon, and especially embodied carbon when it comes to the AEC life cycle. Here we see several stages of a building. It starts all the way on the left with the extraction and transportation of raw materials. Then it continues with the construction phase, followed by the longest phase, which is the operational phase of a building, while the building is being used for its purpose.

And at the end on the right, we have the end of life and the disposal and recycling of the building materials. We refer to the first two stages as embodied carbon. So all the emissions produced in the stages A1 to A5 are referred to as embodied carbon. This is because the carbon that's generated by that time, by the time the building is constructed, we cannot further optimize that carbon. It's wrapped. It's in the materials of the building.

The second type of carbon, let's say, is the operational carbon. These are all the emissions that are produced during the operation of buildings, such as electricity or gas use, but also additional embodied carbon as part of the maintenance and repairs and refurbishment costs and so on. Embodied carbon and operational carbon together are combined into what we call total carbon, and that is the current focus for Autodesk for our impact team.

And if we also take into account the end of life stages, we refer to it as whole life carbon. Having a lot of software for creating BIM models allows us to have a good starting point for creating carbon analysis at all design stages. But let's see how we even calculate this embodied carbon once we have a BIM model. As mentioned, it starts with an annotated BIM model. So we need to know which materials are assigned to which elements.

Then we can extract from the BIM model itself the quantities of these elements and the materials used and create a bill of material or BOM for short. This includes precise masses, volumes, and areas for different elements and materials which we can then match to specific carbon coefficients or carbon factors. And we'll later see where these come from. And as a last step, we can then convert the Bill of Material, magnitude the factors, and calculate the embodied carbon footprint of a particular building or a building model.

So let's now look a little bit closer to a couple of challenges around surrounding the bill of materials. A bill of material of a building is every single component that consists-- that the building consists of. The more components we take into account, the more accurate the embodied carbon estimate will be. The challenge around this is that once we have a BIM model, we can easily extract those values. The BIM model fidelity is very different at very different design stages. It starts with a very rough model, sort of blocky looking building during the conceptual design.

And as we progress through pre-design and construction and pre-construction processes, the architects keep adding more and more details to those buildings. However, if we want-- however, later on we are in the process, the more costly it gets to update the model and improve the carbon efficiency of the building. So we realize we need to create software that can help you assess the carbon of a building early on in the process but also later on. And we have solutions that we'll see later that work with all of these phases.

But now back to the bill of materials. Let's see one challenge around creating a bill of material. So this is a typical wall in Revit that you probably are familiar with. But wall itself is not just a rectangular block. Walls usually have layers. They have insulation, they have plastering, and so on. So how do we account for that when we are creating our estimates?

This is information that's usually not modeled. OK, we could say let's say 40% is structural wall, 10% is rebar, 18% is plaster, and so on. So while this is a good estimate, another thing that's not modeled in the world is how the walls connect the different joints. And if you imagine for, let's say a one family building, this is not a significant difference. But for a multistory building, this can account for several tons of difference of materials.

So it's important to know while we are creating a sustainability analysis, that maybe not information is contained in the models. And that's the first challenge. However, I would like to now focus more on the second portion of this, which is where the carbon coefficients come from and the complexities of this portion of data. Sustainability data is generated by the manufacturers of the materials, manufacturers of products.

It's a very rigorous process, and most commonly this embodied carbon information, as well as other environmental impacts, are contained in a document that's called an environmental product declaration or an EPD for short. These EPDs conform with one of the recognized lifecycle assessment or LCA standards and provide basic information about all the environmental impacts and carbon emissions, acidification, ozone depletion, and so on.

EPDs are specific materials and products, however, there are data sets out there that use EPD data as a basis for creating more generic estimates that are applicable in the early design stages. For the sake of the argument, I will just continue talking about the EPDs. But it's important to realize that similar challenges surround other types of data, and not just in AEC, also in manufacturing, for example, when it comes to digital passports, for example.

The standards are mentioned for EPDs indeed exist. However, they are not technology standards. They are standards for the PDF documents that the manufacturers often have to create and submit as part of their products. And the problem arises when we try to digitalize these types of PDF EPDs because there is no agreed technological standard when it comes to this. As a result, different data sets or different data aggregators end up with slightly different formats of the same data, essentially.

And here we see an example of OKOBAUDAT and EC3. OKOBAUDAT that is a data set that's-- it's a German data set for embodied carbon information that's also used in a wider European space. And EC3 is a data set that's based in the North American space but rapidly expanding to other countries as well. And this is an example of just two of them.

But during my research, I kind of classified these different types of data sets into API based, file based, and proprietary information, proprietary being the custom curated data sets that our users and customers created by the help of other consultancies. But even within these three types, even within the API based data sets, there are differences. Some use JSON formats, some use XML formats, and so on. And even within the JSON based APIs, the data structures are different. EC3 might be using a slightly different data structure as OKOBAUDAT or one click LCA or other API based data sets.

And if you're thinking, oh, there's probably not that many data sets, there are many, many data sets. Here's an example of just a few of the data sets that are applicable in Australia. And I'd like to point out here all of them, even though they contain data for embodied carbon, they even use slightly different methodologies for obtaining the data. And this is just some of the examples where sustainability data gets a little tricky to work with. It's not easily comparable if we want to make fair comparisons.

And the reason that we have so many databases is that all of the data is highly regionalized. It's very applicable only in certain regions. So if I'm building a house in California, I need to use concrete that's probably generated and produced in California. If I'm in UK, I'll probably be using very different manufacturers. As mentioned, different methodologies affect these data sets. The data can be collected and processed and averaged out different ways.

The data can include different lifecycle scopes, as we saw at the beginning. It can include just embodied carbon. It could include some of the predictions for the operational use of the materials, and so on. And as well, when we are generating these types of EPD documents, we need to use certain other data sets for additional calculations. For example, when we are thinking about how much energy was produced while generating and produced-- like manufacturing a certain product, we have to use the electric grid information.

But if electric grid information has bad data, that can significantly affect the embodied carbon of our end product. So I mentioned there's no technology standards. That was maybe badly put. There are some technology standards, however, they're either not widely adopted or not widely agreed. It'd be great if we had just one. However, we identified a few, and they all had their own pros and cons. And you can see them on the screen. They all carry the same information, but the data formats are slightly different. So how can we bring those data sets further together?

So let's pause a little bit. I presented a couple of challenges around generating bills of materials around obtaining sustainability carbon factors, and presented how we need to combine the two to get the embodied carbon estimate. Let's look at what Autodesk is doing to help you with this process. And you may have heard of the Autodesk Platform. We introduced this platform to connect your design and main data and help you achieve your digital transformation goals faster.

The base of this platform is Autodesk Platform Services or APS for short, and it consists of a collection of cloud APIs. An API is an interface between-- an API is the way that software communicates with other software in the same that a UI or user interface is how humans communicate with the software. And APS is a collection of such APIs that all of our products are then essentially built on.

It's also exposed as an external partner-- an external platform, which is trusted and used by thousands of our customers and partners to build their own solutions. By giving you access to the same APIs and services that we are building our entire design and platform on, we're also giving you the key to manage, organize, and connect your enterprise data with the data from our platform, others' products, and third party solutions. With easy access to your data, you can build custom cloud-powered workflows that automate processes, connect applications, and integrate with enterprise systems.

And since we support over 70 file formats and counting, you can rest assured we are committed to supporting bidirectional interoperability and open industry workflows. All of our APIs are built on standard modern web technologies and are constantly evolving and can be grouped in three sets of capabilities. We have the core APIs, which are industry agnostic. Then we have the data APIs, which allow you to interact with the model information, model data in a much more consistent way, in much more efficient way.

And through data exchanges, it allows you to connect your models to external authoring tools as well. And then we have our product APIs, which are designed for you to interact directly with the selected products that we have. These APIs are flexible so they can be leveraged by developers of all skills and levels, whichever programming language they prefer. So you see this platform that is the basis of all our products. It's also used in turn as a basis for all of our sustainability solutions and our current sustainability offering.

And here is an example of several products that we have for sustainability and total carbon analysis when it comes to the AC buildings environment. Due to the evolution of our products, a lot of these solutions are point based solutions. The challenge here is that each solution connects to material data or a sustainability data set, for example, from embodied carbon, with different technologies in different ways.

And also our offering is a mix across the AC products as well as plugins and integrations that link big models to material data. All of these share some of the commonalities that can be streamlined. The opportunity here is to take a platform approach with material data as well as other data to enable the consistency and quality of data in AC. And by building this on the platform, this will also allow us to apply the learnings and extend our capabilities in the manufacturing space as well.

Sustainability data, that you see here on the right-hand side, is country and region specific, as mentioned, and subject to several legislative requirements which result in a range of these non-standardized data sets that our customers often need to rely on. If we want our products here on the left to be applicable and relevant to more of our customers and in more regions, we need to connect them with the data sets they trust, want, and are often required to use. And for this, Autodesk needs to rely on the third party data sets.

For this reason, we are creating a new framework for uniform access to sustainability data for our products as well as our users. It's designed to connect the users with the most relevant data at all design stages. We will achieve this by creating a new API on APS that we call Autodesk Sustainability Data API. It is designed to be extensible beyond just embodied carbon data and is designed to also support all our industries.

Here I want to emphasize that we won't be aggregating the data. Data collection, cleansing, correction, verification, et cetera, is a complex process, and we don't intend to do that ourselves. Instead, the goal of the sustainability data APIs to give the data providers a simpler gateway to get their data into our products and enable our users to use our tools with the data sets they want and need to use.

We will achieve this in two ways. We want to first streamline the extraction of model data for sustainability analysis. This will help our data partners to create the analysis engines in the future. And we also want to create this new API that will create this easy gateway to get third party data into our existing sustainability solutions and future sustainability solutions, creating the opportunity to create that-- which is the opportunity to create that streamlined user experience across the design lifecycle.

Our new APIs will eliminate this integration redundancy by allowing you to only integrate with Autodesk platform once, and then the data will be available to our products but also on the platform. And the way this will work is that at some point we'll have to remap the data into some common data standard. Here we will rely on the latest industry standards, if they exist, or try to provide enough flexibility for everyone to bring their own data into our ecosystem so that our users can work in the most efficient way as they can.

So let's look at what are the benefits of this approach. I want to split the benefits here from three points of view. The first one for the APS developers. These are the people that are already using our APIs or that want to use our APS APIs in the future. By using sustainability data API together with our data model APIs, your applications will be able to more easily switch between different data sources that are applicable only in certain regions, which will automatically extend and broaden the regional applicability of the apps, as well as this also applies to our products, of course.

For the data providers-- these are the companies that own the sustainability data sets. For data providers, this is an opportunity for their data and their data intelligence to reach more users more quickly by just integrating with the APS platform once and then all our products that have and will adopt the sustainability data API will be able to pull that data into their workflows.

And for the end users, the benefit here is that this unlocks this consistent user experience and configurable data sets at each point in their design stage. Ultimately, we want to create this uniform and streamlined workflow powered by our partner data to accelerate sustainable design across the industries.

If this sounds like something you're interested in, whether you're a data set provider, an APS developer, or an end user, we'd love to hear from you. So please scan this QR code or navigate to the URL below the code, and we may contact you about the upcoming feedback sessions or upcoming private and public betas for this API. And we may also keep you up to date about the new developments, so please get in touch.

Of course, it wouldn't be a product demo talk without a product demo. So we'll now take a look at how one of our existing products, ACS Takeoff has integrated a very early prototype of this new sustainability data API and is using it to calculate the embodied carbon of the takeoffs. So first we'll start by creating a new takeoff type, which will be a concrete deck. A concrete deck will consist out of concrete and a couple of metal components.

For each type, we select a unit and we also select a classification. By using a known classification, the takeoff is able to create a call to the sustainability data API and pull in the embodied carbon estimate that is, for this demo, coming from the EC3 data set. As you can see here at the bottom, the estimate is in kilograms of CO2 per ton. In takeoff, you can also adjust this formula to further tweak how embodied carbon is calculated depending on your methodology or your other requirements.

So once you create your takeoff type, you then start mapping out this takeoff component by outlining the 2D elements on this sketch. By doing this, the takeoff will automatically extract the size of your takeoff elements, and from there, the embodied carbon will be calculated. So here we see in the middle of the screen there is an embodied carbon estimate for all the elements that were selected.

And now we can also adjust the values that we receive from EC3. So let's say the sustainability expert in our team decided that we use a different type of concrete. We can adjust this value, and the embodied carbon estimate is automatically recalculated. As we can see here. So this was the demo.

You could see from the challenges that I explained earlier and the complexities when it comes to data and just the vastness of this sustainability space, and also pair that with the urgency of the climate changes, we realize that we cannot build everything on our own. We will continue building our solutions, but at the same time, we want to partner with our customers that are already building additional sustainability solutions available to all our users to further speed up the opportunity for the sustainable design.

We are already working closely with many of our sustainability tech partners who are passionate about sustainability, and you can see them here on this screen. And all of them are willing to build on top of Autodesk capabilities and making it available to all our users. We support our customers and our partners with their strategic business and sustainability goals and help them unlock the true value of Autodesk technology so that they can achieve better outcomes and get to more sustainable data and intelligence in the hands of our customers.

If this sounds like the journey you're on, I'd like to encourage you to consider becoming a sustainability tech partner. And just to name a few of the perks, you will receive exclusive access to our tools. You will get preferential support and training and invitations to events like accelerators. And we also help you promote your solution and your impact story on our social media.

To recap, here's a couple of examples-- here's a couple of ways you can engage with us. First, I'd like to encourage you, if you're interested in contributing to building our sustainability platform and contributing your solutions, to scan the first code to be able to participate in some of our upcoming betas. If you're interested more broadly about the Autodesk platform, you can scan the second code and follow our developer newsletter.

And if you're interested in becoming a partner and building solutions for everyone to use, consider becoming our sustainability tech partner. I hope this gave you a good overview of just the complexity of the sustainability data space and also gave you a good idea of what we are working towards and how we want you to be a part of this story as well. Thank you.