Article
Nexus Pro
min read

Episode #30 reaction: baking your cake with Azure Digital Twins

December 10, 2020

Happy Thursday!

Welcome to this week’s deep dive exclusively for Nexus Pro members. It’s an honor to have you here. This deep dive is a follow up to my recent podcast conversation with Matthew Vogel, Program Manager of Azure IoT Smart Places and Energy Team at Microsoft. I learned a lot from this conversation and want to share my takeaways and the full transcript with you below.

In case you missed it in your inbox, you can find the audio or video here:

Nexus site | Apple Podcasts | Spotify | YouTube | Add to other podcast apps

Enjoy!

—James


Outline

  • My reaction, including highlights
  • Full transcript

My reaction

This is your last podcast reaction, my friends. Starting in 2021, we’re going to kill this weekly deep dive routine and provide you wonderful members weekly value in a different way. If you’re wondering why, it’s because we can see that barely any of you read these things. 👀  👀  👀

Okay, so now that I know 70+% of you don’t like the term digital twin, I’m going to react to this episode by using it 27 times. 😬

I liked Matt’s two-folded digital twin definition. First, it’s replicate the physical world with the following pillars:

  • Live data

  • Structure

    • Ontology - codifying the industry expertise into a model

    • Topology - entities that use those models

  • Geometry and/or visualization - provides end-user experience

  • Behavior - live execution environment

  • History - look back and learn

  • Connection to your lines of business - CRM, CMMS, etc

Then, to Microsoft, it’s also their platform play: they provide the base ingredients that are tailor-made for parts of that definition above and provide the connections between them. You can be the baker for your customers, they’re going to stock your pantry with baking stuff. The set of capabilities and APIs that helps you create the structure and connect it with the other Azure IoT services.

I know you don’t like the term digital twin, but ignore this episode at your own risk, my friends. Matt is talking about some new marketplace dynamics we haven’t seen before from the building owner’s perspective - reducing their risk so they can switch AND forcing vendors to conform to the data model. Allowing vendors to plug and play. Allowing owners to start with one use cases, like space management, and then use the same infrastructure to add on new use cases. Teaching building owners to procure smart building products differently: infrastructure separate from applications. These trends, if they take off, will create nonlinear effects where the best solutions start winning out faster and faster.

Finally, we have another competing data model - two weeks in a row! Sorry. Here are my notes on what Matt said about DTDL…

  • DTDL is the underlying modeling language - not an ontology itself - just a schema for names, properties, and relationships
  • Then you layer that with sets of models on top of DTDL and that’s where RealEstateCore comes in
  • Matt says RealEstateCore was chosen because when it’s used in practice, it’s used as is

My highlights:

  • Matthew answers James’ favorite question - it’s a different skillset and different set of stakeholders (3:33)
  • interaction between Matthew’s team on the product side and the Redmond campus project (6:31)
  • Defining digital twins and clearing up common misconceptions (9:37)
  • Relationship between Azure and digital twin partners like Willow (13:39)
  • How their approach enables portability among vendors (29:03)
  • How building owners should get started - one use case at a time, or build out the digital twin and see where the data takes you? ; prioritization of use cases based on ROI (35:22)
  • Where DTDL and RealEstateCore fit in the greater industry context (40:41)
  • What Matthew is excited about - accelerating time to results (51:46)

Full transcript

Note: transcript was created using an imperfect machine learning tool and lightly edited by a human (so you can get the gist). Please forgive errors!

James Dice: [00:00:00] Hello, friends. Welcome to Nexus, a smart buildings technology podcast for smart humans. I'm your host, James Dice. If we haven't met before, I write a weekly newsletter on the same topic. It's also called Nexus. Each week I share what I've learned, my opinions, and what I'm excited about in the quickly evolving world of intelligent buildings. Readers have called Nexus the best way to stay up to date on the future of this industry without all the marketing fluff. You can check it out and subscribe at nexus.substack.com or click the link in the show notes.

Since starting the Nexus newsletter, many of you have reached out to me wanting to talk shop, and we have. After a few weeks of those wonderful conversations, I realized I needed to record and share them with our growing community. So here we are. The Nexus podcast is born. This is our chance to explore and learn with the brightest in our industry together.

Episode 30 is a conversation with Matthew Vogel of Microsoft's Azure IOT smart places in energy team. I met Matthew a few weeks ago when he sat on a panel, I moderated at real calm, which was great fun. If you're wondering what Microsoft is up to in the smart building space. This episode is definitely for you. We talked about Microsoft's Azure, digital twin platform and ecosystem, and how it, as Matthew says, is designed to accelerate the time to results for the smart buildings market. We talk about how they're doing that and why i see value in it of course, we also covered the data modeling aspect of the digital twin, including the open source digital twin definition language, and where that sits in context with similar efforts we've covered on other episodes, please.

All right. Hello, Matt. Welcome to the nexus podcast. Can you introduce yourself for us?

Matthew Vogel: [00:01:54] Yeah, thanks for having me. Um, uh, Matthew Bogle, I'm a program manager on our Azure IOT smart places and energy team at Microsoft.

so I've been at Microsoft for around eight and a half years now spending across, uh, office mobiles. I helped launch the initial versions across windows, phone, Android, iPhone. Uh, I did some early experimentation with IOT and office around natural language, which led me to a team, that had started in Skype for IOT, for communications.

Uh, so I helped launch our company. Kardon invoke our poor Thomas speaker, uh, with the state domain, as well as starting our partnership with Amazon to get segued into echo devices. and then over the last three plus years, I've been at dryer team in a few different functions, but all focused on smart buildings and commercial real estate.

James Dice: [00:02:38] Got it. Got it. and what was it like going from like this general it space into buildings?

Matthew Vogel: [00:02:45] Like how did,

James Dice: [00:02:46] how, how was that transition for you?

Matthew Vogel: [00:02:48] Yeah, it's been very interesting. I think on the consumer side, things tend to move a little bit faster and so there's a lot more innovation upfront, but I think the buildings industry has kind of seen what's happened in the consumer space and tried to start adopting it.

And so you saw some early versions of that, but we were. but you definitely see digital transformation taking hold in buildings. And, for me, it's been interesting to just learn about a completely different industry, compared to productivity or communications, or had been in for five plus years.

James Dice: [00:03:18] Nice. Yeah. for those of you who've listened to the episode with the manual, Daniel, what would that have been like a month or two ago? This will be kind of like Microsoft part two, uh, digging into the product side of Microsoft. And that's where Matt, really comes in strong here.

So I want to start us off by getting to my favorite question, which is, I've been asking this on the last, I don't know. 12 or so podcasts episodes. So we have a nice sort of repository of, of different answers and what's fun is they're all different. So, why do you think technology and the, like you just said, why do you think technology and buildings is sort of like decades behind the technology in our pockets or in the consumer space?

Sure.

Matthew Vogel: [00:03:58] Um, I think the first thing that comes to mind is that a lot of the, the new wave of technology, innovation, things like artificial intelligence cloud, you know, especially now digital twins, it's a very different technology focus than the traditional technology focuses the in. So the traditional more hardware, you know, maintenance management system type of thing.

And so part of it is just being a completely new skillset for people. Um, but part of it too, with, you know, I think is the general theme of digital transformation is making sure that you have all of the right stakeholders involved and it's changing the way you implement technology. It's changing the way that you do business.

I think a lot of this has to change, but it's affecting the way many people across the ecosystem are doing business today.

James Dice: [00:04:44] Yeah. And that is a unique answer. I like that because. I don't know if that is fully grasped. Right. And I'm teaching this course on smart buildings right now. And I, think when people, when the students, I see the students kind of start to grasp this concept where it's like, we're not just talking about technology here, we're talking about a new way of doing business.

And that suddenly sort of like balloons the topic. Right. But, but you have to start asking these sorts of questions for you to have like an impact with the technology. Right.

Matthew Vogel: [00:05:15] Right. Yeah. And a microcosm of that for example, is when we started selling the surface hubs. So the 70, 80, 90 inch, you know, PCs, Microsoft initially went into it with a traditional it focus.

We thought it's just like selling any other PC where you just have to sell to the it department. And you're good to go. But because you're now attached to these things to walls, which means you need to run more power, you need to run other networking. You need to run all of your accessories, your audio video equipment to that machine, not even about the facilities people.

And so it's a whole different set of stakeholders that have their own agenda and their own priorities, their own requirements too, in terms of network security and infrastructure. So all of that has to plan together and they're not always in sync. So it becomes that. That issue of where you have to start addressing, how do you get these people to align on potentially conflicting goals or conflicting outcomes?

James Dice: [00:06:09] Totally. Yeah. So this is where I kind of want to tie this back to the episode with the manual. So manual, and I talked about basically the Redmond campus and the transformation happening there with the setup. I think new buildings. So he talked about building smart from the ground up.

Right. And I thought it was brilliant. So he walked us through like, The entire stack, of sort of what's happening on campus there. And what I wanted to ask you about was how has your team, and first of all, maybe you could tell us, like what a little bit about your team,

but how has your team on the product side

at Microsoft sort of learned from the journey that, Emmanuel's team has gone through, at the Redmond campus?

Matthew Vogel: [00:06:47] Sure. Yeah, I guess first for the lay of the land of how Emmanuel's team and my team worked together, Uh, Emmanuel's team is a broader team called applied innovation where part of their team is focused directly on our own campus to ensure that we're handed the right digital transformation. And like you mentioned, making sure that tech and smarts and all these use cases are built in from the ground up.

but they also help enable partners and customers in the space with their own digital transformation externally. Uh, they bring a lot of great contacts and leads from the field. And for my team, as the product group were responsible for making sure that our own products let's say IOT hub or digital twins or computer vision, any of these other pieces of product that you see across Microsoft, making sure that they could be tailor made to what you need to use in the industry, as well as coming up with the right design patterns to show how all of these Lego bricks can fit together to enable a holistic solution for the industry.

and campus is a great way of showing how that comes together, because we've decided from the start, the campus needs to be a showcase, not just the smart building technology and envision, but something that has to be repeatable externally. If we're building something on our own campus that someone else can't come in and visit and say, yeah, I want to adapt that right away to my building, my portfolio.

if they can't do that with whatever partners we brought to campus or whatever technology we built on campus, then I think ultimately we fell together because can't just be Microsoft in a bubble. Um, it's our job to enable the ecosystem at large. And that's how we all went together. and personally, I like to equate it to something like the surface program where, you know, PCs have kind of stagnated a little in terms of form factors and stuff.

So, we let out with, you know, surface to say, this is a new two in one format with a detachable keyboard and kickstand. then eventually that was adopted by the broader PC industry. Um, there are ways that we could do that now with our own campus where we showcase here's the latest that you can do and show that as something that the ecosystem can adapt.

But at the same time, there are learnings from the field we could bring in. And so you saw that with, let's say the two screen phones and stuff now where, you know, we now have that a surface duo I'd imagine that the same way. If we're seeing innovation happen externally with other partners or customers of the landlords, we should also be willing to kind of be humble about it and adopt that into our campus and learning how to grow our own campus that way too.

Got

James Dice: [00:09:11] it. Yeah. But it works both ways for sure. All right. So I want to kind of zero in a little bit on, so you guys, when we say Microsoft products, there's a lot of products, right? I kind of want to zero in for this conversation a little bit more around digital twins. Uh, so you and I were on a panel at real com a couple of weeks ago together, which was super fun.

and. I want to sort of maybe replay a little bit of that conversation for those of the audience that didn't listen to it or didn't watch it.

So

how do you guys define digital twins? And, the second piece of that is like, what are the misconceptions around it that you're continuously clearing up when you're talking about digital

twins?

Matthew Vogel: [00:09:48] Sure. Yeah. So, uh, digital twins itself is not something we invented yet. You know, a concept that's been around for decades now, especially with, I think NASA is responsible for creating it for assimilating. A lot of that physical world in a digital environment. Um, for us at Microsoft, I think we have two different versions of a definition just to keep things complicated.

But, uh, I'd say the general, the view we have is that a holistic digital twin solution is something that can replicate the physical environment in the digital world. But pillars of that holistic solution include things like live data, whether it be from IOT, data sources, other devices, third-party sources like weather or traffic mobility, uh, you need live data feeding in.

you need some form of structure and infrastructure. We talk about ontologies, which is your vernacular, your way of, codifying your industry expertise and your knowledge into a set of models. as well as your typology, which is your live instance of entities that use those models. So, think of like a central of a conference room where a conference room has a capacity and, and properties that metadata like temperature and, how many people could hold it.

So on. the instance of that in your topology would be conference in one, two, three, which is located on floor three of building 40. and that can hold eight people in currently it's 70 degrees. and so that's part of the structure aspect of it. Uh, we talk about geometry or visualization because ultimately you need to deliver it some sort of end experience to your users.

And so in a building that's typically a floor plan or a bin model. Um, we talk about, behavior. So why have execution environments that way, if something goes wrong in, let's say part of your, HVAC system, you know, exactly which zones that impacted, which rooms are in those zones and then which people are allocated to those rooms.

you need some out of history, so you can look back and learn from, how your environment has adjusted. So, whether that be data over time or even the structure of your graph over time. And then lastly is connection to lines of business. So whether that would be a CRM tool, it could be a work order management tool.

It could be your productivity apps, a digital twin solution needs, right. That connection to other data sources or other end points as well. and so for us at Microsoft, we talk about a digital twin solution encompassing all of that. That's where a lot of partners who I'm sure we'll talk about later on, come into play to stitch it all together because the second definition I'll give you is again, we are a platform company.

Uh, we provide the base ingredients, the capabilities that have to be assembled. and so for us, Azure, digital twins is our product that's tailored, made towards specific parts of that solution. So our agile digital twins, platform is a set of API APIs and capabilities that helps you create an ontology for your domain.

Uh, instantiate it topology. And then it connected with some of these other services like IOT hub, which is that IOT data source teller, or Azure maps creator, which is the indoor maps part of the world or time series insights, which is more of a story. And so, on the you know, micro level, we think of digital twins as our platform, as a service offering that empowers more holistic, solutions by partners and customers.

James Dice: [00:13:03] Hmm. All right. There's a lot there. so  this is the number one reason I wanted to get you on the show because like when I hear digital twin and when I hear you describe that, I think we should unpack it a little bit more because I think when people hear, say, Willow, talk about their digital twin offering.

Maybe just use that as an example. So Willow provides a digital twin that sits on top of Azure, digital

twins. And I think that concept is very

difficult to say, like, where does. The Azure pay stop. And where does Willow begin and vice versa. Right. So how would you sort of demystify that aspect and Willow is just one example of, the partners on the marketplace,

Matthew Vogel: [00:13:47] right?

Yeah. And the analogy I like to use, actually, I'm trying to take people out of the tech world for this one where okay. I'd talk about baking in a way where Microsoft has a platform company is responsible for creating all the ingredients that you'd see in your pantry. Think about that, supplying your flour, your sugar, or your eggs, or solves whatever else you want to do and what to use.

Um, but we rely on partners like Willow, who you mentioned, but a number of others that, we've published and talked about, uh, externally. they're the bakers who come to our pantry to assemble, to bake a cake or to make bread or to, you know, I guess the amount of bread, a good pandemic recipe to kickoff, but yeah,

James Dice: [00:14:29] a lot of people are doing sound.

Matthew Vogel: [00:14:31] Yeah. So they they're the ones with, development expertise that can look at this entrance ingredients and bake a cake that makes sense for their customers. In the case of Willow, they build something that's very operations focused. So they. Take our ingredients to, uh, import a BIM file, turn it into a graph with digital twins, and then create an entire user experience around that tailor made to people who are operating your building.

you could change your, your perspective and then look at someone like Steelcase. Uh, Steelcase's a furniture manufacturer, but they also designed office spaces and they actually were the inception of digital twins for us because they came to us. You know what the problem statement that they wanted to understand how their space is being used.

They help design these amazing spaces, but they needed data to prove that that was adding value or, or how that space is actually being occupied day over day. Um, and so for them, their digital twin based solution is something that can not only visualize your space, but show you how often something's being used.

And what factors might play into that to help the designers or silly managers. Understand how they should be configuring their space. Okay. Um, back to one of your earlier questions about a misconception of a digital twin, is it doesn't need to be one thing. Everyone will have their own unique take on a digital twin and that's absolutely.

Okay. For us at Microsoft, that's our job to make sure our pantry of ingredients can be customizable enough, to make all those different recipes that our customers might want to create. but then, we need to make sure that we offer an ecosystem of partners that have these interesting, uh, solutions or capabilities for smart buildings, industry.

That way they can come together to a holistic smart building, because we know that a smart building isn't just operations. Isn't just space utilization. Isn't just energy. all of that has to come together into one cohesive, deployment. Okay.

James Dice: [00:16:25] So The partner provider or the provider, like Willow or sort of Steelcase they're coming in and they're saying, I'm going to stitch all these or combine all these ingredients together to bake our loaf of bread.

Right.

Matthew Vogel: [00:16:38] And that

James Dice: [00:16:39] loaf of bread is then what's provided to end user. Right. And so you guys are just saying, you know, which of our ingredients would you like to use essentially? And you have many, many, many different facets of it Is the ontology and the taxonomy is that piece. So you mentioned when you were describing your definition or your pillars, you mentioned that piece being sort of, they can create their own, like these, the bakers themselves can create their own ontology taxonomy.

Right. But then there's the efforts to sort of open that piece up and standardize it as well. Right. So. How does that show up? And how is that a piece of your platform that you're providing? If that,

Matthew Vogel: [00:17:23] yeah. Um, I'll, I'll see how far we can stretch this analogy, but, uh, basically, I think first, the, the first part of that is around accelerating time to results.

if we, as Microsoft have to train every single Baker that comes into our kitchen to understand how to bake and understand how these ingredients come together. Then we're losing time. And in turn, the customers are losing time to value because everyone is and learning how to add yeast and flour and all that stuff.

Yeah. They're to, you know, make their bread. if we, as Microsoft can provide a starter kit, an accelerant to your development, then that's a win for everyone. So the ontology you mentioned, it's something we're specifically working with real estate for, uh, they're a consortium based out of Europe where.

they bring a lot of academic experience with semantic web and apology development, and they created this real estate court model, that is used directly in the real world today. They have a variety of customers and partners actually building real world solutions on top of that oncology. And so what we've done is we've partnered with them to actually provide an open source version of real estate core on top of digital twins definition, language.

which is the underlying modeling language that you would use, for your Azure digital twins deployment. and so if we kind of bring back the ontology that we were talking about before, as part of agile, digital twins, basically, do you need to start agile, digital twins with a set of models. So we know how to parse the information that you're sending on us.

and we don't want people looking at a blank page and figuring out how to start with those models, because we've already proven it out of what works in the industry. So we get to leverage a lot of the real estate core expertise and proven value. Bring that to the digital twins platform and an open source way that everyone can leverage.

So you're getting a quicker time to results because you're not having to create everything from scratch. Uh, Willow was one of those key collaborators on the models, but also one of the key consumers of it. And they've actually talked about the fact that they've saved hundreds of developer hours just in the span of a month.

because they weren't spending time creating their own models, on their own. Uh, so that's one aspect of it. Uh, the second aspect that we were trying to address with a lot of the, the standardization that you talked about is, when we were talking about all those different solutions that you have to fit together into a building.

what we heard from property owners today is that it's all proprietary. It's all walled gardens, and it's really hard for them to actually break those barriers down and draw conclusions or insights together. So for example, if Steelcase were to tell you that this room is occupied or not, how do you then go back to the building management system and control the HVAC?

based on that occupancy right now, you're dealing with proprietary formats. you know, in any side of the world, or if they do open formats, then it's hard to integrate those together. And so what we're proving out now is the fact that if more and more of these solution providers, the three on those same, ontologies on those same models, it becomes easier for the integrator in the middle for that holistic digital twin, the building to bring that all together.

So again, kind of break us out of technology. The analogy I like to use there is. if let's say you were the building owner and I came to you and I was selling you a solution that talked about with bathrooms and another solution provider came in and sold you a solution that talked about your washrooms and someone else came in and talked about your restrooms.

We were all talking about the exact same rooms in the exact same building, but you're sitting above that thinking I'm getting all this different data talking about these different rooms. How do I make sense of what's what and how do I join that depth together? Because ultimately the physical space is still the same.

by using that, not only the same language, so digital twins, definition language, but also the same dialect of that language that apology the real estate core models. now you can actually provide a connected ecosystem of partners, vendors, and customers to use those same languages, to write better integration through Azure digital twins.

James Dice: [00:21:17] So, so D TDL digital twin definition language is. Sort of like a non domain specific

language. And

then what you do is layer the domain specific or the real estate specific

ontology on top of that.

Right. And that's what you've sort of said. I think what I'm hearing too is that, that real estate specific one.

Matthew Vogel: [00:21:37] any partner could like

James Dice: [00:21:39] basically develop their own on top of DT DL. But what you're saying is why don't we sort of go to the building owners and say, you guys need to standardize on your domain specific ontology. That way all of these providers can then easily work together.

Matthew Vogel: [00:21:54] Exactly. Yeah. It's another great example of how our campus and how our, partner is actually starting to put together to where on our campus.

We're actually starting to apply that ontology too. the solutions that we're onboarding, how we do RFPs, how we actually choose vendors, because we know that the ones using that ontology will be that much easier for us to integrate into our campus. now we actually have other property owners, other managers who are looking at it and saying, Hey, I'm dealing with the same issue that you're dealing with on your campus, but they're dealing at a portfolio level.

they can now do the exact same thing where they required the real estate court ontology. For any of their RFPs or for any of their solution finders. that's where we have a number of different partners throughout the stack actually started to build that. And, it's great to see that you're kind of, we address it from both sides.

Now you have address it from the bottom up with the supply of solutions that are building on that platform. You're also addressing it top down because now you have the property owners and managers who are proving that they haven't needed value for that ontology. Got

James Dice: [00:22:53] it. Got it. So I'm still sort of piecing all this together in my mind.

So one piece that I don't know that I understand at this point is at the edge. So I'm going to go out there and I'm going to deploy some sort of driver to talk to each individual system in a traditional building. You have all these disconnected systems. So how does

that

play into the overall Azure digital twin stack?

And do you guys have partners that are just, all they're doing is mapping from that original protocol into this new, sort of open,

Matthew Vogel: [00:23:23] ontology? Sure. Yeah, absolutely. Um, so there, there are a couple aspects there. one cool part about digital twins, definition, language DDL is that it's not just an abstract, modeling language for space or for all these other entities.

There's actually a device centric view of that, that we call IOT plug and play, which is how you actually model the devices using that same language. So for newer devices, for the Greenfield, you can actually, as a device manufacturer, use that exact same language and still apply that same ontology to your motion sensor, to your, you know, CIC beam to your, HVAC systems, whatever else you want to model.

but at the same time we realized that the scale in the industry is still in the brownfield. It's still the legacy buildings and all of the instrumentation and systems that you're never going to touch again. And like you mentioned, machine learning and, you know, translation becomes really important that scenario.

so we're working with a number of partners, you know, some like Optio three, for example, who apply machine learning on the edge, to. Discover classify and normalize the legacy data. You're getting into the destination of your choice. And so real estate Corp obviously becomes that destination that they're training towards.

and that also becomes an accelerant to brownfield onboarding. we know that today it typically takes a month or, you know, even up to three or more months to onboard a building. And when you're dealing with portfolio managers who manage hundreds of thousands of buildings, but the physics just doesn't work, you know, people will be out in three years by the time you onboard all their buildings.

and so we know we need to shrink that time down to, you know, days, if not hours. and automate that process as much as possible. And so that's where we do believe in the value of having a machine learning on the edge to help with that translation and adapt from, the Bradfield devices into something more standardized and well known.

Yeah. So

James Dice: [00:25:15] when we're talking about an ecosystem, it could either be a device manufacturer that says, Hey, I'm. IOT plug and play certified or whatever is that I don't even know if

Matthew Vogel: [00:25:23] that's the right way to put it.

James Dice: [00:25:26] When you have an integrator that says I can, you know, enable this data to be displayed in the right way or, enable some sort of API to happen that has the data model in the right way.

Matthew Vogel: [00:25:36] Right. And it keeps your integration layer very thin because now you're no longer building a lot of that custom integration code to speak from solution to device. Everyone's agreed on the same format. So as long as that contract is divided by, then it'll just kind of seamlessly work. Got it.

James Dice: [00:25:52] Okay, cool.

So now that we sort of. Have that picture painted. Hopefully it's painted well for, people on what we're talking about here, but let's talk about sort of. How that enables a smarter building or how that enables buildings to be smarter, worldwide.

Right? How are we pushing the industry forward with this platform, with this ecosystem approach?

and I thought I

wanted to circle back around to something you said when we were prepping for the real calm conversation where I think you said you started with Steelcase. And you had a certain approach to developing the platform, and then you said you did it, and then you guys took a step back and said, well, wait, or we kind of need to rethink how we're doing this.

Can you kind of explain that story?

Matthew Vogel: [00:26:34] Yeah. So, like I mentioned earlier to, you know, Steelcase was, an early partnership with us that started the core of the idea that we needed the association between the people, places and devices. and then gradually as we grew that, we recognized the ability for those, functions of capabilities that the relationships.

To apply to other solutions and other scenarios. And so that's how that one partnership turned into a broader platform. in our first version of Azure digital twins that we launched in 2018, that was very focused on the concept of people, places, and devices. Those are actually, types that were baked into the platform by default.

And then you can only kind of tweak it somewhat. we still found that that had tremendous value and it changed the way people thought about building solutions in the industry. But a lot of the feedback we received was the fact that, those three concepts don't necessarily apply directly to every single solution or even if it does.

You still want the ability to model it more, customize it a little bit more than what we offer. Okay. Um, and so in the next wave of digital twins that we updated a few months ago, as well as now, generally about the last October, digital twins definition, language DTL became part of, or a big part of the platform capabilities.

and that opened us up to a number of other industries, whether it be manufacturing vehicles, even, retail and so on, or the energy grid. those can now all model their own domain expertise, on DTL and build tailor-made solutions for what they need. so it helped us break out, uh, the smart buildings industry and become much more broadly applicable.

and then as we identify other needs that are coming from these other markets, then that's where we could identify. What needs to be additional capabilities in the platform, uh, what needs to be a solution accelerator? So whether it's the ontology or the connection between digital twins and natural paths for visualization, how do we identify those?

And then lastly, what are the types of partners, customers who could leverage those capabilities and that's where we've grown our ecosystem for the real estate industry with. You know, people who go across the stack for operations, construction, engineering, you know, space occupancy returned to work, productivity experiences, and so on.

James Dice: [00:28:49] Got it. That's really cool. That's it amazing progression starting out with buildings and then saying, you know, where else can we go? Normally you see companies try to come in to the building space after being successful elsewhere. Interesting. So we described this whole, this ecosystem approach where there's a bunch of different ingredients, as you said.

So if I'm a building owner like a large REIT, for instance, If I know that my solution providers are building on top of this ecosystem, how does that enable me to, have portability between vendors for instance? Cause I think that's a, an underestimated way that.

This space is being held back because people are scared to invest. If they know that they kind of have to get married to a vendor. And they're kind of scarred from being married to certain vendors for the past several decades. Right.

Matthew Vogel: [00:29:40] So a lot of baggage in the industry.

James Dice: [00:29:42] Yeah. So much baggage. Yeah. so how does this approach sort of help that?

I don't know if it, it does, but I'm, I'm feeling like it would make it so that you could maybe start to plug and play vendors if you wanted to.

Matthew Vogel: [00:29:55] Exactly. Yeah. I think that's really a part of it too, is that, let's say, uh, I've decided that I want partner X for one of my solutions. Let's say building management system.

I installed partner X and partner X. Isn't living up to it, but. Because I've required part next to be honest, real estate core ontology. I could easily swap out there, you know, my building management system, or, you know, some other solution with partners, Y or Z or whoever else can also buy it by that.

So I no longer have to change my core infrastructure just because I've swapped someone out, because I've mandated that anything coming into my building uses this language uses this ontology. Um, the other aspect of where I think digital twins helps a property owner or manager is nuggets are drawing insights between these solutions.

and so you can start saying all of these disparate silos can now start influencing each other. That could be from, the digital twin that the building owner manages, and the digital twin that the tenant managers. So, uh, typically, you know, a building owner will have a blind spot for a tenant space and a tenant won't necessarily have access to all the shared building infrastructure.

having that common language, having that integration layer, but digital twin can provide you the right connections between them, or just even between other solutions in your own domain. So I mentioned before the occupancy and the building management system, or it can even think it would say elevators and a calendaring system, you just start understanding where an elevator needs to go because of, you know, the meetings that are on the calendar that day.

or your work order management. I have you start understanding who's in what space, when our space is clear, what our space is being utilized and how do you prioritize which issues are resolved based on the severity of the issue who it's impacting, where it's impacting, or what systems is impacting, having that twin can start allowing you to draw context from, pieces of all these different solutions together.

Got

James Dice: [00:31:51] it. I love that. Yeah. And there's a couple, yeah. Different threads that have come up in several of the past couple of podcasts, which is one, is like our state-of-the-art right now, if you just look at the buildings industry as a whole, our state-of-the-art is really, the point solution  or the siloed application.

Right. So you're, you're starting to get to where that is kind of becoming a thing of the past. Right? Um, the other, like aspect of this as the. Easier for an owner to sort of fire a vendor. It is right there.

That's a double-edged sword. So the vendor might not like that, but that same setup allows them to also integrate really fast.

Right. So they can just plug their application directly into this. Existing stack, whereas before they had to do a ton of work and you mentioned people doing the same work, like our industry is plagued by everyone sort of doing the same shit. Right? Uh, everybody's riding the same drivers. Everyone's doing the same.

Same integration. Everyone's, adding their own ontologies off of the same point lists like we're doing so much, that's the same. And so the double-edged sword is like, yes, you can get fired tomorrow. If your SAS product doesn't work. But also you can also plug it into way more, peoples.

Buildings. So I think that's going to be an accelerator in that the best solutions will sort of win out faster. Right. And

Matthew Vogel: [00:33:11] that's an important industry for if we come back. Uh, I remember the first, one of the first questions you're asking me is why is digital transformation, this industry so slow? I think that's a great example of it.

People are used to in this industry, like they know how to run RFP processes. They know how to choose a vendor based on the individual merits of that vendor. You can look at five different building managers and systems or, you know, space utilization systems and you know, and the traditional software checkbox that does have this feature does that this feature do I like the UX and isolation, that makes sense.

And, and the Industry is able to do that. but now it's a very different type of question you have to ask them as part of that process is how does this accrued to my digital twin infrastructure? and what are the outcomes I want to draw? What are the insights I want to take from that solution in conjunction with everything else that I have in my building, the very different way of looking at that problem space, because now you have to manage not just the vendor itself, but also be integration.

Yeah. and so I think that's learnings that we're seeing and that's what people are adapting to right now. but kind of come back to your last statement too. we are seeing a lot more momentum for the people who are jumping on this early, because there is a sort of first mover advantage this from a property owner perspective where, the property owners that are embracing digital twins and this type of integration model are clearly seeing, not just cost savings, but also value adds.

and enough, so that we have partners in the space who are demonstrably saying, I can show that a building with a digital twin is more valuable, an asset than a non digital twin building. so they're seeing it from that side, but at the same time, the vendors who are actually starting to transform their solutions to use these ontologies and to use digital twins, are that much more accepted by these property owners who are mandating it as part of their digital infrastructure.

so we, we do see it taking off very quickly. We see, you know, exponential growth in that area. and it's kind of among the first times we're actually seeing the two ends really meet in the middle to drive the ecosystem and the industry at large.

James Dice: [00:35:13] Got it. You just reminded me of something that I wanted to circle back to from our conversation at real calm.

So. Well, something that came up there was like, how does the building owner gets started? Right. So you mentioned the perspective of a very large building owner and they're coming at it from like, I'm going to go tell all my asset managers that we're going to value buildings differently, but it's something digital twin and like, okay.

That's pretty progressive from, from that standpoint. But like, If I want to get started

tomorrow, the value that makes that asset worth more is I'm implementing all of these use cases, right? I'm touching the occupants, I'm hitting tenants, that's a big undertaking, right.

To go from where we're at today, which is point solutions. At best two, like this fully enabled digital twin experience. Right. So one of the things we talked about and the panel discussion was, do you start with one use case, prove that out and grow it out, or do you start with, you know, just getting all the data into the twin and then it will become clear later on where the value is at.

So I want you to sort of repeat your answer or you can come up with a better one, uh, from what we talked about on the panel,

Matthew Vogel: [00:36:21] Yeah, think it's a huge issue that, people face when they're dealing with digital twins is it's very easy to get lost in the art of the possible, or even just the academia of a digital twin and spend just hours talking about it and the possibilities and all that.

And the possibilities are true and the possibilities are, I guess we could call it endless. That might be cliche, but, the key, to actually unlocking that value is getting started somewhere. and you know, early on in the process for us since we were new to the real estate industry as a company, we engaged a lot with our corporate strategy team to understand how we should interface with the industry and what role we had to play in.

And we actually had a similar discussion, a lot of the world's largest landlords as well. and so we quickly gravitated towards that three 3,300 model that everyone talks about in the industry. So, companies spend $3 per square foot per year on energy, 30 on space, 300 out of the people in the space.

And naturally it's Microsoft and we've seen other companies do the same thing. We jumped on that 300 number. We said, that's the biggest number on the slide. We said, we're the most experienced there because of our background office. let's go tackle that. we quickly realized number one, that's where you get very lost in all the possibilities.

But number two, the value is a little bit squishier there. we know that there is value in saving people time and then making people more comfortable and helping people be more effective. But do you quantify that in salary? Do you quantify that? And you know, time-saver, you know, there's a bunch of variables that aren't necessarily easy to pull up the count that it says, save this much money, or I added this much value.

Whereas with the three number, you could very easily say I'm turning off this light switch, or I am turning off my HVAC, or I'm doing this other thing to save energy. I know exactly how much money I'm saving there. It's a very quantifiable and what we've actually seen on our own campuses. We can actually break even on that within 12 to 18 months.

And so at that point afterwards, you're printing money for your, implementation and you can just keep growing from there. the next piece of that would space is still quantifiable. You know how much, least you have, you know, how much real estate you're owning. it's just a little bit less liquid.

So it's a little harder to impact directly because you might have contracts or you're not going to take a sledgehammer to your wall every day. but we know that there is value in understanding how space is being used and how you can reconfigure it. Especially now, given a lot of the changes with, uh, COVID pandemic.

and so the interesting part of that though, is that it's not a separate solution necessarily. It could build upon the foundation. You've set with your investments in the energy and operation space, because eventually what you're going to change about your space, but also impact the way you use energy.

So again, back to my example of, you can turn off the lights and turn off the heating. If you know that space is typically not used during given hours, they're automating that and start optimizing. Then eventually when you do get to the people in the space, you're also still going to need to know what the spaces itself, the design, the capacity, all of that.

You're still gonna need to know how to interact with the building management system and the controls system in your building. But now that you've added those concrete players in the three and the 30 space, you can more easily tackle the 300. So the, partners and the customers that we've seen be more successful in this space are able to very clearly define what their business goals are.

how they want to achieve that and also lay out a very clear plan about how they're going to go tackle it, because it, I think people should dream about the different possibilities. They have three, five, 10 years down the line, but if you don't get mobilized on something realistic and something concrete in the short term, you're just going to get lost.

In theory for years. so we've seen people basically like RXR, for example, carved out a lab within their headquarters. So they could start implementing there first and start experimenting. But then they had a very crisp vision about how they start rolling out from that one lab to their broader building, to a set of buildings, to their entire portfolio.

that's a very great example of how you achieve success at portfolio scale, by starting with something very crisp and painting a roadmap about how to evolve.

James Dice: [00:40:20] Got it. I think that's a great answer on like letting ROI sort of drive your, Prioritization of your use cases. And then like the ROI really depends. This is what I say in the course, ROI really depends on how fluffy are willing to accept the fluffiness factor of what you're willing to accept.

And so I think that's a great point. so let's transition to sort of, I want to circle back to DTL because. obviously this is not the only, effort to create some sort of ontology happening in the industry. And so one of the things that I'm trying to do is sort of clarify all of these different efforts, and sort of not really compare them to each other, but say where they fit in the context of the other things that are going on.

Right. So I've done several podcasts now around. haystack brick. Um, the previous one, when this, when this show airs, the previous one will have been Google and their, digital buildings ontology. so we talked about that. Um, so want you to kinda talk about where DTL fits and then obviously real estate Corp.

That's sitting on top of DTL Like we talked about earlier, um, where that fits in this. Grand scheme of things of, of similar efforts.

Matthew Vogel: [00:41:30] Sure. Yeah. So DTL TDL, I'd say it's the underlying modeling language. So it's actually not an ontology per se itself. but it's just a scheme by which you could describe any entity as well as the relationships between them.

So it's just, it's basically it's Jason LD format, and it's just meant to say, this is a structured way of you, describing the name, the capabilities, the properties, the telemetry of your entities and the connections between them. that's actually all open source. It's not something proprietary that we want to own.

It's just a way of us saying this is a structured way of, the ability for us to bring data into Azure, digital twins, such that you can model any entity you want. as a digital twins in turn has parsers and capabilities to take DTL and make something of it. But there's nothing stopping people externally from looking at DTL and making their own parsers or toolkits around their own platforms to inherit that as well.

Yeah. then you start layering that up with ontologies or sets of models on top of digital twins, definition, language, which is where real estate core comes in for us. we spent a lot of time trying to figure out do we come up with our own ontology? Do we inherit from a few ontologies? what do we do there?

And. For us, it was extremely important, not to reinvent the wheel both because we didn't want to spend our own time, you know, doing that when there's obviously a lot of learnings in the industry. also because we didn't want to be like the XKCD comic where there's. Yeah, there's one slide. There's 15 competing standards.

I'll build a new standard that addresses all of them. And then fast forward, there is 16 competing centers. Um, we didn't want that to happen. We wanted to leverage what was there already. And, and real estate Corp was something that was provably, adding value to the industry in a way that, was actually used in practice in the real world.

so the reason why we went forward with real estate Corp, not just for the merits of the ontology itself. Was also because when it's used in practice, it's used as is, uh, when we looked at things like brick and haystack, people tend to have their own flavor of it. Your version of brick might be very different from my version of brick and it would be really hard for that property manager to integrate both of those together.

Totally. With real estate Corp, they actually have an existing set of partners and customers who are using the ontology as it. So one of the examples I gave in my I become blog was, I knew which builds up prop tech O S solution, and some of their key customers are people like Basset Cronin, which is one of the biggest, construction and property managers in Europe.

There's Y it, which is one of the biggest facility managers in Finland, and they also operate across Northern Europe. and so they consume that solution and they consume it using the models as they are described in the ontology itself. So we, we no longer have that dialect issue because, it's used kind of to the letter of the models.

Cool.

James Dice: [00:44:17] And, and real estate course something that's not as well-known in the U S it doesn't come up a whole lot. Right. So, it might make sense to sort of give a little bit more background on how that relates to, like you mentioned, it's used as is, but how is it able to be used as is, and is it, an open source project that's happening just elsewhere that people can contribute as well?

Matthew Vogel: [00:44:39] Exactly. Yeah. So it's open source. They have a kind of consortium that looks at the contributions and make sure that, the models kind of adhere to the overall connections and the overall, Voice and kind of tone of all the different models and how they fit together. and then what they did very purposely from the start was to make sure that they had developers and customers lined up to use those models directly.

So they have, uh, a really interesting blend of academic stakeholders, technology stakeholders, and real estate stakeholders, all of whom can validate, the ontology end to end. Um, they, they came to us because they're interested in digital sprints a long time ago. but we were always intrigued by the idea of an industry leading ontology.

And so as we held the capabilities to, open up DTL to allow any modeling, that's where it became much more possible for us to go back to them and say, Let's actually build the next version of real estate Corp using DTL. Uh, and so now you can actually use the DTL language, to leverage real estate core, but then your actual digital twin solution.

So there's nothing stopping people from still using real estate Corp, as it was previously built in their own solutions, but now we make it that much easier to connect it into anything at your IOT based because we'll know how to kind of adhere to that TTL contract.

James Dice: [00:45:56] Got it. I love it. So when I

read your blog, by the way, and we'll put your blog in the show notes, I want to circle a one sentence in it, which said.

Real estate Corp specifically does not aim to be a new standard, but rather provides a common denominator of bridges, other standards, such as brick project haystack and more. So is it like translating between

different ontologies or how does

that, what does that mean?

Matthew Vogel: [00:46:18] Yeah. So, when they first started out, they actually do take a lot yeah.

Concepts from things like break in haystacks that way an owl, as well as another language that they can bring in. So they do have an ability to translate models externally into what they use today. and so that was also an interesting concept that they, you could have a bridge between, ontologies to help with your on ramp and accelerate your time to results again.

but I think where it's also, more interesting for us is that, once you're in that ontology, you could use it directly in the tooling that you have. and you know, in our version of the world, its DTL and our partners and the national digital twins. and so again, it was kind of a one-off translation.

We did to make sure that real estate court could be built with DTL. But now what we've proven is that you no longer have to, translate or convert. You could just kind of start with that tooling as it exists in the repo today. at the same time, we do recognize that there might be blind spots that we have.

There might be gaps. There might be areas where we need people to bring in their industry expertise. and so what we worked on with the real estate core group was ensuring the plugability of the ontologies that way. Let's say if we don't have elevators, so we don't have deep expertise on elevators, you know, apartments like disapprove elevator or, others might be better suited to describe the entire model of an elevator in the systems and data could send you, um, that's where we allow them to contribute their models to extend, but make sure that it's not Island.

It's not, they're not a siloed ontology. It fits together within the right asset classes or equipment or within the broader building structure.

James Dice: [00:47:53] Got it makes total sense. So how are you seeing, I know you guys have a lot of projects going on right now, or how are you seeing a building owner say, I want to standardize on this

ontology.

Like how do I, as a building owner, make sure that this happens in my building.

Matthew Vogel: [00:48:09] Yeah. So, I think, again, it starts with the business value. You're trying to drive because even though the intelligi is very complete, it's pretty rare for anyone to use an entire ontology, in any given solution.

You're going to take a subset of that ontology for the. Pieces of the domain. That makes sense for you. So the parts of the ontology, you use her as space management solution. It's very different than the ones that you'll use for an operations, solution. so the first thing is making sure you have your clear view of the business goals, but then you can start kind of driving conversations with partners and customers about.

what are the solutions you can provide and how do you plug into my infrastructure? So a lot of the bigger property managers who've seen. So if we come back to the logging in, I mentioned Brookfield and Oxford, they basically say they want to drive the holistic digital twin infrastructure of their buildings.

First that way, they have the right backbone to kind of hang these other solutions onto. and they basically say that in that solution, we want, our ontology, our domain expertise to be real estate core ontology, because moving forward as they engage with specific partners or specific solution providers in the industry, they can say that here is the structure I have in my building.

Here's how I will allow you to integrate with it. how can you go plug in with me? Uh, in some cases, we find that the opportunity is big enough for a solution provider to completely rearchitect their solution, to use the ontology natively in their software as a service model. In other cases, we've seen, vendors, whether it just as a kind of quick and dirty implementation or desperate time or resources sake.

provide a translator on eat grass. So that way, even if the entire city, the solution isn't rearchitected, you can still translate into the data that your consumer expects the data format to be in.

James Dice: [00:49:55] Got it, this whole, platform aspect and the value of it. I've struggled to integrate it into my course, because the course is introductory.

And I want to sort of paint the picture of like where the industry is at today, but it's almost like I need a separate module. That's like, we're used to integrating and creating these point solutions that are sort of full stack. there are building owners out there that are just going the platform route and saying, okay, now

your solution will

integrate back to me.

Uh, and I haven't quite figured out how to, deploy that in the course. So yeah. That's advice. Anybody has advice. Let me know,

Matthew Vogel: [00:50:29] uh, as we. Good analogy I've also seen happen is like take PCs today. We'll take for granted the fact that you could plug in a keyboard or plugging a mouse or plugging a monitor, and it just works under the covers.

That's actually happening is some sort of plug and play type logic where your operating system has dictated. This is the contract by which I will recognize your equipment. If you say that I am a monitor, I am a keyboard or I am a mouse. And these are my capabilities. I have a scroll wheel. I have to click buttons.

I have assigned click bullet. I don't know exactly how to pull in your drivers. I don't know exactly how to command it. Control your device. Now you take that same concept into a building and you have the building owner saying, this is the operating system in my building. it could be a Willow, it could be a Bentley, it could be magic.

You know, any of those players who are targeting the holistic twin of a building, then you could say, that's my contract. That's my integration layer. Here's how you can go play in this ecosystem. And so now you can start driving more of those solution providers to better plug and play with your building.

James Dice: [00:51:32] Totally. I love that. Cool. So as we kind of close things out, I wanted to ask you and, I put you on the spot here. What are you sort of excited about? I know you guys just made several announcements, but if you look a couple of years out, uh, what are you excited about on the Microsoft product side of things?

Matthew Vogel: [00:51:49] Um, I think one key thing is making sure we keep accelerating time to results. So I think now We've shown that there's value here in that ontology and in the ability to plug and play these solutions into the digital twin. the magic has to be in shrinking the time to results from again, months into days or hours.

We need that to be as quick and seamless as possible for any of this to drive, at scale value. we've seen now it's, not easy, but at least there is line of sight and straightforward for anyone in the industry to go tackle one building at a time with spot solutions. Yeah. Uh, for someone to succeed, it's going to take thousands of buildings and showing that you can do this at portfolio scale and minutes or hours or maybe days.

that's one thing that I'm really excited about. And then the other piece that's exciting for me is. from a Microsoft perspective, we're seeing all of these worlds, from our, industry collide together. So Microsoft is unique in that it has that three cloud story. It has, Azure, it has an Azure IOT.

It has dynamics and it has office. This industry is perfect for that intersection. And so I do see. a lot more of that value coming to property owners, tenants, uh, workplace solution providers, and so on, to leverage all of the capabilities from Microsoft, for any given solution to the industry.

James Dice: [00:53:09] Well, cool. I appreciate you coming on the show. I think this'll be a great follow on to Emmanuel's episode, but also a lot of the other ones as well. So, thanks so much and we'll talk to you soon.

Matthew Vogel: [00:53:20] Yeah. Thanks for having me.

James Dice: [00:53:21] Alright, friends. Thanks for listening to this episode of the Nexus podcast. For more episodes like this and to get the weekly Nexus newsletter, please subscribe at nexus.substack.com. You can find the show notes for this conversation there as well. As always, please reach out on LinkedIn with any thoughts on this episode.

I'd love to hear from you. Have a great day

Upgrade to Nexus Pro to continue reading

Upgrade

Upgrade to Nexus Pro to continue reading

Upgrade

Happy Thursday!

Welcome to this week’s deep dive exclusively for Nexus Pro members. It’s an honor to have you here. This deep dive is a follow up to my recent podcast conversation with Matthew Vogel, Program Manager of Azure IoT Smart Places and Energy Team at Microsoft. I learned a lot from this conversation and want to share my takeaways and the full transcript with you below.

In case you missed it in your inbox, you can find the audio or video here:

Nexus site | Apple Podcasts | Spotify | YouTube | Add to other podcast apps

Enjoy!

—James


Outline

  • My reaction, including highlights
  • Full transcript

My reaction

This is your last podcast reaction, my friends. Starting in 2021, we’re going to kill this weekly deep dive routine and provide you wonderful members weekly value in a different way. If you’re wondering why, it’s because we can see that barely any of you read these things. 👀  👀  👀

Okay, so now that I know 70+% of you don’t like the term digital twin, I’m going to react to this episode by using it 27 times. 😬

I liked Matt’s two-folded digital twin definition. First, it’s replicate the physical world with the following pillars:

  • Live data

  • Structure

    • Ontology - codifying the industry expertise into a model

    • Topology - entities that use those models

  • Geometry and/or visualization - provides end-user experience

  • Behavior - live execution environment

  • History - look back and learn

  • Connection to your lines of business - CRM, CMMS, etc

Then, to Microsoft, it’s also their platform play: they provide the base ingredients that are tailor-made for parts of that definition above and provide the connections between them. You can be the baker for your customers, they’re going to stock your pantry with baking stuff. The set of capabilities and APIs that helps you create the structure and connect it with the other Azure IoT services.

I know you don’t like the term digital twin, but ignore this episode at your own risk, my friends. Matt is talking about some new marketplace dynamics we haven’t seen before from the building owner’s perspective - reducing their risk so they can switch AND forcing vendors to conform to the data model. Allowing vendors to plug and play. Allowing owners to start with one use cases, like space management, and then use the same infrastructure to add on new use cases. Teaching building owners to procure smart building products differently: infrastructure separate from applications. These trends, if they take off, will create nonlinear effects where the best solutions start winning out faster and faster.

Finally, we have another competing data model - two weeks in a row! Sorry. Here are my notes on what Matt said about DTDL…

  • DTDL is the underlying modeling language - not an ontology itself - just a schema for names, properties, and relationships
  • Then you layer that with sets of models on top of DTDL and that’s where RealEstateCore comes in
  • Matt says RealEstateCore was chosen because when it’s used in practice, it’s used as is

My highlights:

  • Matthew answers James’ favorite question - it’s a different skillset and different set of stakeholders (3:33)
  • interaction between Matthew’s team on the product side and the Redmond campus project (6:31)
  • Defining digital twins and clearing up common misconceptions (9:37)
  • Relationship between Azure and digital twin partners like Willow (13:39)
  • How their approach enables portability among vendors (29:03)
  • How building owners should get started - one use case at a time, or build out the digital twin and see where the data takes you? ; prioritization of use cases based on ROI (35:22)
  • Where DTDL and RealEstateCore fit in the greater industry context (40:41)
  • What Matthew is excited about - accelerating time to results (51:46)

Full transcript

Note: transcript was created using an imperfect machine learning tool and lightly edited by a human (so you can get the gist). Please forgive errors!

James Dice: [00:00:00] Hello, friends. Welcome to Nexus, a smart buildings technology podcast for smart humans. I'm your host, James Dice. If we haven't met before, I write a weekly newsletter on the same topic. It's also called Nexus. Each week I share what I've learned, my opinions, and what I'm excited about in the quickly evolving world of intelligent buildings. Readers have called Nexus the best way to stay up to date on the future of this industry without all the marketing fluff. You can check it out and subscribe at nexus.substack.com or click the link in the show notes.

Since starting the Nexus newsletter, many of you have reached out to me wanting to talk shop, and we have. After a few weeks of those wonderful conversations, I realized I needed to record and share them with our growing community. So here we are. The Nexus podcast is born. This is our chance to explore and learn with the brightest in our industry together.

Episode 30 is a conversation with Matthew Vogel of Microsoft's Azure IOT smart places in energy team. I met Matthew a few weeks ago when he sat on a panel, I moderated at real calm, which was great fun. If you're wondering what Microsoft is up to in the smart building space. This episode is definitely for you. We talked about Microsoft's Azure, digital twin platform and ecosystem, and how it, as Matthew says, is designed to accelerate the time to results for the smart buildings market. We talk about how they're doing that and why i see value in it of course, we also covered the data modeling aspect of the digital twin, including the open source digital twin definition language, and where that sits in context with similar efforts we've covered on other episodes, please.

All right. Hello, Matt. Welcome to the nexus podcast. Can you introduce yourself for us?

Matthew Vogel: [00:01:54] Yeah, thanks for having me. Um, uh, Matthew Bogle, I'm a program manager on our Azure IOT smart places and energy team at Microsoft.

so I've been at Microsoft for around eight and a half years now spending across, uh, office mobiles. I helped launch the initial versions across windows, phone, Android, iPhone. Uh, I did some early experimentation with IOT and office around natural language, which led me to a team, that had started in Skype for IOT, for communications.

Uh, so I helped launch our company. Kardon invoke our poor Thomas speaker, uh, with the state domain, as well as starting our partnership with Amazon to get segued into echo devices. and then over the last three plus years, I've been at dryer team in a few different functions, but all focused on smart buildings and commercial real estate.

James Dice: [00:02:38] Got it. Got it. and what was it like going from like this general it space into buildings?

Matthew Vogel: [00:02:45] Like how did,

James Dice: [00:02:46] how, how was that transition for you?

Matthew Vogel: [00:02:48] Yeah, it's been very interesting. I think on the consumer side, things tend to move a little bit faster and so there's a lot more innovation upfront, but I think the buildings industry has kind of seen what's happened in the consumer space and tried to start adopting it.

And so you saw some early versions of that, but we were. but you definitely see digital transformation taking hold in buildings. And, for me, it's been interesting to just learn about a completely different industry, compared to productivity or communications, or had been in for five plus years.

James Dice: [00:03:18] Nice. Yeah. for those of you who've listened to the episode with the manual, Daniel, what would that have been like a month or two ago? This will be kind of like Microsoft part two, uh, digging into the product side of Microsoft. And that's where Matt, really comes in strong here.

So I want to start us off by getting to my favorite question, which is, I've been asking this on the last, I don't know. 12 or so podcasts episodes. So we have a nice sort of repository of, of different answers and what's fun is they're all different. So, why do you think technology and the, like you just said, why do you think technology and buildings is sort of like decades behind the technology in our pockets or in the consumer space?

Sure.

Matthew Vogel: [00:03:58] Um, I think the first thing that comes to mind is that a lot of the, the new wave of technology, innovation, things like artificial intelligence cloud, you know, especially now digital twins, it's a very different technology focus than the traditional technology focuses the in. So the traditional more hardware, you know, maintenance management system type of thing.

And so part of it is just being a completely new skillset for people. Um, but part of it too, with, you know, I think is the general theme of digital transformation is making sure that you have all of the right stakeholders involved and it's changing the way you implement technology. It's changing the way that you do business.

I think a lot of this has to change, but it's affecting the way many people across the ecosystem are doing business today.

James Dice: [00:04:44] Yeah. And that is a unique answer. I like that because. I don't know if that is fully grasped. Right. And I'm teaching this course on smart buildings right now. And I, think when people, when the students, I see the students kind of start to grasp this concept where it's like, we're not just talking about technology here, we're talking about a new way of doing business.

And that suddenly sort of like balloons the topic. Right. But, but you have to start asking these sorts of questions for you to have like an impact with the technology. Right.

Matthew Vogel: [00:05:15] Right. Yeah. And a microcosm of that for example, is when we started selling the surface hubs. So the 70, 80, 90 inch, you know, PCs, Microsoft initially went into it with a traditional it focus.

We thought it's just like selling any other PC where you just have to sell to the it department. And you're good to go. But because you're now attached to these things to walls, which means you need to run more power, you need to run other networking. You need to run all of your accessories, your audio video equipment to that machine, not even about the facilities people.

And so it's a whole different set of stakeholders that have their own agenda and their own priorities, their own requirements too, in terms of network security and infrastructure. So all of that has to plan together and they're not always in sync. So it becomes that. That issue of where you have to start addressing, how do you get these people to align on potentially conflicting goals or conflicting outcomes?

James Dice: [00:06:09] Totally. Yeah. So this is where I kind of want to tie this back to the episode with the manual. So manual, and I talked about basically the Redmond campus and the transformation happening there with the setup. I think new buildings. So he talked about building smart from the ground up.

Right. And I thought it was brilliant. So he walked us through like, The entire stack, of sort of what's happening on campus there. And what I wanted to ask you about was how has your team, and first of all, maybe you could tell us, like what a little bit about your team,

but how has your team on the product side

at Microsoft sort of learned from the journey that, Emmanuel's team has gone through, at the Redmond campus?

Matthew Vogel: [00:06:47] Sure. Yeah, I guess first for the lay of the land of how Emmanuel's team and my team worked together, Uh, Emmanuel's team is a broader team called applied innovation where part of their team is focused directly on our own campus to ensure that we're handed the right digital transformation. And like you mentioned, making sure that tech and smarts and all these use cases are built in from the ground up.

but they also help enable partners and customers in the space with their own digital transformation externally. Uh, they bring a lot of great contacts and leads from the field. And for my team, as the product group were responsible for making sure that our own products let's say IOT hub or digital twins or computer vision, any of these other pieces of product that you see across Microsoft, making sure that they could be tailor made to what you need to use in the industry, as well as coming up with the right design patterns to show how all of these Lego bricks can fit together to enable a holistic solution for the industry.

and campus is a great way of showing how that comes together, because we've decided from the start, the campus needs to be a showcase, not just the smart building technology and envision, but something that has to be repeatable externally. If we're building something on our own campus that someone else can't come in and visit and say, yeah, I want to adapt that right away to my building, my portfolio.

if they can't do that with whatever partners we brought to campus or whatever technology we built on campus, then I think ultimately we fell together because can't just be Microsoft in a bubble. Um, it's our job to enable the ecosystem at large. And that's how we all went together. and personally, I like to equate it to something like the surface program where, you know, PCs have kind of stagnated a little in terms of form factors and stuff.

So, we let out with, you know, surface to say, this is a new two in one format with a detachable keyboard and kickstand. then eventually that was adopted by the broader PC industry. Um, there are ways that we could do that now with our own campus where we showcase here's the latest that you can do and show that as something that the ecosystem can adapt.

But at the same time, there are learnings from the field we could bring in. And so you saw that with, let's say the two screen phones and stuff now where, you know, we now have that a surface duo I'd imagine that the same way. If we're seeing innovation happen externally with other partners or customers of the landlords, we should also be willing to kind of be humble about it and adopt that into our campus and learning how to grow our own campus that way too.

Got

James Dice: [00:09:11] it. Yeah. But it works both ways for sure. All right. So I want to kind of zero in a little bit on, so you guys, when we say Microsoft products, there's a lot of products, right? I kind of want to zero in for this conversation a little bit more around digital twins. Uh, so you and I were on a panel at real com a couple of weeks ago together, which was super fun.

and. I want to sort of maybe replay a little bit of that conversation for those of the audience that didn't listen to it or didn't watch it.

So

how do you guys define digital twins? And, the second piece of that is like, what are the misconceptions around it that you're continuously clearing up when you're talking about digital

twins?

Matthew Vogel: [00:09:48] Sure. Yeah. So, uh, digital twins itself is not something we invented yet. You know, a concept that's been around for decades now, especially with, I think NASA is responsible for creating it for assimilating. A lot of that physical world in a digital environment. Um, for us at Microsoft, I think we have two different versions of a definition just to keep things complicated.

But, uh, I'd say the general, the view we have is that a holistic digital twin solution is something that can replicate the physical environment in the digital world. But pillars of that holistic solution include things like live data, whether it be from IOT, data sources, other devices, third-party sources like weather or traffic mobility, uh, you need live data feeding in.

you need some form of structure and infrastructure. We talk about ontologies, which is your vernacular, your way of, codifying your industry expertise and your knowledge into a set of models. as well as your typology, which is your live instance of entities that use those models. So, think of like a central of a conference room where a conference room has a capacity and, and properties that metadata like temperature and, how many people could hold it.

So on. the instance of that in your topology would be conference in one, two, three, which is located on floor three of building 40. and that can hold eight people in currently it's 70 degrees. and so that's part of the structure aspect of it. Uh, we talk about geometry or visualization because ultimately you need to deliver it some sort of end experience to your users.

And so in a building that's typically a floor plan or a bin model. Um, we talk about, behavior. So why have execution environments that way, if something goes wrong in, let's say part of your, HVAC system, you know, exactly which zones that impacted, which rooms are in those zones and then which people are allocated to those rooms.

you need some out of history, so you can look back and learn from, how your environment has adjusted. So, whether that be data over time or even the structure of your graph over time. And then lastly is connection to lines of business. So whether that would be a CRM tool, it could be a work order management tool.

It could be your productivity apps, a digital twin solution needs, right. That connection to other data sources or other end points as well. and so for us at Microsoft, we talk about a digital twin solution encompassing all of that. That's where a lot of partners who I'm sure we'll talk about later on, come into play to stitch it all together because the second definition I'll give you is again, we are a platform company.

Uh, we provide the base ingredients, the capabilities that have to be assembled. and so for us, Azure, digital twins is our product that's tailored, made towards specific parts of that solution. So our agile digital twins, platform is a set of API APIs and capabilities that helps you create an ontology for your domain.

Uh, instantiate it topology. And then it connected with some of these other services like IOT hub, which is that IOT data source teller, or Azure maps creator, which is the indoor maps part of the world or time series insights, which is more of a story. And so, on the you know, micro level, we think of digital twins as our platform, as a service offering that empowers more holistic, solutions by partners and customers.

James Dice: [00:13:03] Hmm. All right. There's a lot there. so  this is the number one reason I wanted to get you on the show because like when I hear digital twin and when I hear you describe that, I think we should unpack it a little bit more because I think when people hear, say, Willow, talk about their digital twin offering.

Maybe just use that as an example. So Willow provides a digital twin that sits on top of Azure, digital

twins. And I think that concept is very

difficult to say, like, where does. The Azure pay stop. And where does Willow begin and vice versa. Right. So how would you sort of demystify that aspect and Willow is just one example of, the partners on the marketplace,

Matthew Vogel: [00:13:47] right?

Yeah. And the analogy I like to use, actually, I'm trying to take people out of the tech world for this one where okay. I'd talk about baking in a way where Microsoft has a platform company is responsible for creating all the ingredients that you'd see in your pantry. Think about that, supplying your flour, your sugar, or your eggs, or solves whatever else you want to do and what to use.

Um, but we rely on partners like Willow, who you mentioned, but a number of others that, we've published and talked about, uh, externally. they're the bakers who come to our pantry to assemble, to bake a cake or to make bread or to, you know, I guess the amount of bread, a good pandemic recipe to kickoff, but yeah,

James Dice: [00:14:29] a lot of people are doing sound.

Matthew Vogel: [00:14:31] Yeah. So they they're the ones with, development expertise that can look at this entrance ingredients and bake a cake that makes sense for their customers. In the case of Willow, they build something that's very operations focused. So they. Take our ingredients to, uh, import a BIM file, turn it into a graph with digital twins, and then create an entire user experience around that tailor made to people who are operating your building.

you could change your, your perspective and then look at someone like Steelcase. Uh, Steelcase's a furniture manufacturer, but they also designed office spaces and they actually were the inception of digital twins for us because they came to us. You know what the problem statement that they wanted to understand how their space is being used.

They help design these amazing spaces, but they needed data to prove that that was adding value or, or how that space is actually being occupied day over day. Um, and so for them, their digital twin based solution is something that can not only visualize your space, but show you how often something's being used.

And what factors might play into that to help the designers or silly managers. Understand how they should be configuring their space. Okay. Um, back to one of your earlier questions about a misconception of a digital twin, is it doesn't need to be one thing. Everyone will have their own unique take on a digital twin and that's absolutely.

Okay. For us at Microsoft, that's our job to make sure our pantry of ingredients can be customizable enough, to make all those different recipes that our customers might want to create. but then, we need to make sure that we offer an ecosystem of partners that have these interesting, uh, solutions or capabilities for smart buildings, industry.

That way they can come together to a holistic smart building, because we know that a smart building isn't just operations. Isn't just space utilization. Isn't just energy. all of that has to come together into one cohesive, deployment. Okay.

James Dice: [00:16:25] So The partner provider or the provider, like Willow or sort of Steelcase they're coming in and they're saying, I'm going to stitch all these or combine all these ingredients together to bake our loaf of bread.

Right.

Matthew Vogel: [00:16:38] And that

James Dice: [00:16:39] loaf of bread is then what's provided to end user. Right. And so you guys are just saying, you know, which of our ingredients would you like to use essentially? And you have many, many, many different facets of it Is the ontology and the taxonomy is that piece. So you mentioned when you were describing your definition or your pillars, you mentioned that piece being sort of, they can create their own, like these, the bakers themselves can create their own ontology taxonomy.

Right. But then there's the efforts to sort of open that piece up and standardize it as well. Right. So. How does that show up? And how is that a piece of your platform that you're providing? If that,

Matthew Vogel: [00:17:23] yeah. Um, I'll, I'll see how far we can stretch this analogy, but, uh, basically, I think first, the, the first part of that is around accelerating time to results.

if we, as Microsoft have to train every single Baker that comes into our kitchen to understand how to bake and understand how these ingredients come together. Then we're losing time. And in turn, the customers are losing time to value because everyone is and learning how to add yeast and flour and all that stuff.

Yeah. They're to, you know, make their bread. if we, as Microsoft can provide a starter kit, an accelerant to your development, then that's a win for everyone. So the ontology you mentioned, it's something we're specifically working with real estate for, uh, they're a consortium based out of Europe where.

they bring a lot of academic experience with semantic web and apology development, and they created this real estate court model, that is used directly in the real world today. They have a variety of customers and partners actually building real world solutions on top of that oncology. And so what we've done is we've partnered with them to actually provide an open source version of real estate core on top of digital twins definition, language.

which is the underlying modeling language that you would use, for your Azure digital twins deployment. and so if we kind of bring back the ontology that we were talking about before, as part of agile, digital twins, basically, do you need to start agile, digital twins with a set of models. So we know how to parse the information that you're sending on us.

and we don't want people looking at a blank page and figuring out how to start with those models, because we've already proven it out of what works in the industry. So we get to leverage a lot of the real estate core expertise and proven value. Bring that to the digital twins platform and an open source way that everyone can leverage.

So you're getting a quicker time to results because you're not having to create everything from scratch. Uh, Willow was one of those key collaborators on the models, but also one of the key consumers of it. And they've actually talked about the fact that they've saved hundreds of developer hours just in the span of a month.

because they weren't spending time creating their own models, on their own. Uh, so that's one aspect of it. Uh, the second aspect that we were trying to address with a lot of the, the standardization that you talked about is, when we were talking about all those different solutions that you have to fit together into a building.

what we heard from property owners today is that it's all proprietary. It's all walled gardens, and it's really hard for them to actually break those barriers down and draw conclusions or insights together. So for example, if Steelcase were to tell you that this room is occupied or not, how do you then go back to the building management system and control the HVAC?

based on that occupancy right now, you're dealing with proprietary formats. you know, in any side of the world, or if they do open formats, then it's hard to integrate those together. And so what we're proving out now is the fact that if more and more of these solution providers, the three on those same, ontologies on those same models, it becomes easier for the integrator in the middle for that holistic digital twin, the building to bring that all together.

So again, kind of break us out of technology. The analogy I like to use there is. if let's say you were the building owner and I came to you and I was selling you a solution that talked about with bathrooms and another solution provider came in and sold you a solution that talked about your washrooms and someone else came in and talked about your restrooms.

We were all talking about the exact same rooms in the exact same building, but you're sitting above that thinking I'm getting all this different data talking about these different rooms. How do I make sense of what's what and how do I join that depth together? Because ultimately the physical space is still the same.

by using that, not only the same language, so digital twins, definition language, but also the same dialect of that language that apology the real estate core models. now you can actually provide a connected ecosystem of partners, vendors, and customers to use those same languages, to write better integration through Azure digital twins.

James Dice: [00:21:17] So, so D TDL digital twin definition language is. Sort of like a non domain specific

language. And

then what you do is layer the domain specific or the real estate specific

ontology on top of that.

Right. And that's what you've sort of said. I think what I'm hearing too is that, that real estate specific one.

Matthew Vogel: [00:21:37] any partner could like

James Dice: [00:21:39] basically develop their own on top of DT DL. But what you're saying is why don't we sort of go to the building owners and say, you guys need to standardize on your domain specific ontology. That way all of these providers can then easily work together.

Matthew Vogel: [00:21:54] Exactly. Yeah. It's another great example of how our campus and how our, partner is actually starting to put together to where on our campus.

We're actually starting to apply that ontology too. the solutions that we're onboarding, how we do RFPs, how we actually choose vendors, because we know that the ones using that ontology will be that much easier for us to integrate into our campus. now we actually have other property owners, other managers who are looking at it and saying, Hey, I'm dealing with the same issue that you're dealing with on your campus, but they're dealing at a portfolio level.

they can now do the exact same thing where they required the real estate court ontology. For any of their RFPs or for any of their solution finders. that's where we have a number of different partners throughout the stack actually started to build that. And, it's great to see that you're kind of, we address it from both sides.

Now you have address it from the bottom up with the supply of solutions that are building on that platform. You're also addressing it top down because now you have the property owners and managers who are proving that they haven't needed value for that ontology. Got

James Dice: [00:22:53] it. Got it. So I'm still sort of piecing all this together in my mind.

So one piece that I don't know that I understand at this point is at the edge. So I'm going to go out there and I'm going to deploy some sort of driver to talk to each individual system in a traditional building. You have all these disconnected systems. So how does

that

play into the overall Azure digital twin stack?

And do you guys have partners that are just, all they're doing is mapping from that original protocol into this new, sort of open,

Matthew Vogel: [00:23:23] ontology? Sure. Yeah, absolutely. Um, so there, there are a couple aspects there. one cool part about digital twins, definition, language DDL is that it's not just an abstract, modeling language for space or for all these other entities.

There's actually a device centric view of that, that we call IOT plug and play, which is how you actually model the devices using that same language. So for newer devices, for the Greenfield, you can actually, as a device manufacturer, use that exact same language and still apply that same ontology to your motion sensor, to your, you know, CIC beam to your, HVAC systems, whatever else you want to model.

but at the same time we realized that the scale in the industry is still in the brownfield. It's still the legacy buildings and all of the instrumentation and systems that you're never going to touch again. And like you mentioned, machine learning and, you know, translation becomes really important that scenario.

so we're working with a number of partners, you know, some like Optio three, for example, who apply machine learning on the edge, to. Discover classify and normalize the legacy data. You're getting into the destination of your choice. And so real estate Corp obviously becomes that destination that they're training towards.

and that also becomes an accelerant to brownfield onboarding. we know that today it typically takes a month or, you know, even up to three or more months to onboard a building. And when you're dealing with portfolio managers who manage hundreds of thousands of buildings, but the physics just doesn't work, you know, people will be out in three years by the time you onboard all their buildings.

and so we know we need to shrink that time down to, you know, days, if not hours. and automate that process as much as possible. And so that's where we do believe in the value of having a machine learning on the edge to help with that translation and adapt from, the Bradfield devices into something more standardized and well known.

Yeah. So

James Dice: [00:25:15] when we're talking about an ecosystem, it could either be a device manufacturer that says, Hey, I'm. IOT plug and play certified or whatever is that I don't even know if

Matthew Vogel: [00:25:23] that's the right way to put it.

James Dice: [00:25:26] When you have an integrator that says I can, you know, enable this data to be displayed in the right way or, enable some sort of API to happen that has the data model in the right way.

Matthew Vogel: [00:25:36] Right. And it keeps your integration layer very thin because now you're no longer building a lot of that custom integration code to speak from solution to device. Everyone's agreed on the same format. So as long as that contract is divided by, then it'll just kind of seamlessly work. Got it.

James Dice: [00:25:52] Okay, cool.

So now that we sort of. Have that picture painted. Hopefully it's painted well for, people on what we're talking about here, but let's talk about sort of. How that enables a smarter building or how that enables buildings to be smarter, worldwide.

Right? How are we pushing the industry forward with this platform, with this ecosystem approach?

and I thought I

wanted to circle back around to something you said when we were prepping for the real calm conversation where I think you said you started with Steelcase. And you had a certain approach to developing the platform, and then you said you did it, and then you guys took a step back and said, well, wait, or we kind of need to rethink how we're doing this.

Can you kind of explain that story?

Matthew Vogel: [00:26:34] Yeah. So, like I mentioned earlier to, you know, Steelcase was, an early partnership with us that started the core of the idea that we needed the association between the people, places and devices. and then gradually as we grew that, we recognized the ability for those, functions of capabilities that the relationships.

To apply to other solutions and other scenarios. And so that's how that one partnership turned into a broader platform. in our first version of Azure digital twins that we launched in 2018, that was very focused on the concept of people, places, and devices. Those are actually, types that were baked into the platform by default.

And then you can only kind of tweak it somewhat. we still found that that had tremendous value and it changed the way people thought about building solutions in the industry. But a lot of the feedback we received was the fact that, those three concepts don't necessarily apply directly to every single solution or even if it does.

You still want the ability to model it more, customize it a little bit more than what we offer. Okay. Um, and so in the next wave of digital twins that we updated a few months ago, as well as now, generally about the last October, digital twins definition, language DTL became part of, or a big part of the platform capabilities.

and that opened us up to a number of other industries, whether it be manufacturing vehicles, even, retail and so on, or the energy grid. those can now all model their own domain expertise, on DTL and build tailor-made solutions for what they need. so it helped us break out, uh, the smart buildings industry and become much more broadly applicable.

and then as we identify other needs that are coming from these other markets, then that's where we could identify. What needs to be additional capabilities in the platform, uh, what needs to be a solution accelerator? So whether it's the ontology or the connection between digital twins and natural paths for visualization, how do we identify those?

And then lastly, what are the types of partners, customers who could leverage those capabilities and that's where we've grown our ecosystem for the real estate industry with. You know, people who go across the stack for operations, construction, engineering, you know, space occupancy returned to work, productivity experiences, and so on.

James Dice: [00:28:49] Got it. That's really cool. That's it amazing progression starting out with buildings and then saying, you know, where else can we go? Normally you see companies try to come in to the building space after being successful elsewhere. Interesting. So we described this whole, this ecosystem approach where there's a bunch of different ingredients, as you said.

So if I'm a building owner like a large REIT, for instance, If I know that my solution providers are building on top of this ecosystem, how does that enable me to, have portability between vendors for instance? Cause I think that's a, an underestimated way that.

This space is being held back because people are scared to invest. If they know that they kind of have to get married to a vendor. And they're kind of scarred from being married to certain vendors for the past several decades. Right.

Matthew Vogel: [00:29:40] So a lot of baggage in the industry.

James Dice: [00:29:42] Yeah. So much baggage. Yeah. so how does this approach sort of help that?

I don't know if it, it does, but I'm, I'm feeling like it would make it so that you could maybe start to plug and play vendors if you wanted to.

Matthew Vogel: [00:29:55] Exactly. Yeah. I think that's really a part of it too, is that, let's say, uh, I've decided that I want partner X for one of my solutions. Let's say building management system.

I installed partner X and partner X. Isn't living up to it, but. Because I've required part next to be honest, real estate core ontology. I could easily swap out there, you know, my building management system, or, you know, some other solution with partners, Y or Z or whoever else can also buy it by that.

So I no longer have to change my core infrastructure just because I've swapped someone out, because I've mandated that anything coming into my building uses this language uses this ontology. Um, the other aspect of where I think digital twins helps a property owner or manager is nuggets are drawing insights between these solutions.

and so you can start saying all of these disparate silos can now start influencing each other. That could be from, the digital twin that the building owner manages, and the digital twin that the tenant managers. So, uh, typically, you know, a building owner will have a blind spot for a tenant space and a tenant won't necessarily have access to all the shared building infrastructure.

having that common language, having that integration layer, but digital twin can provide you the right connections between them, or just even between other solutions in your own domain. So I mentioned before the occupancy and the building management system, or it can even think it would say elevators and a calendaring system, you just start understanding where an elevator needs to go because of, you know, the meetings that are on the calendar that day.

or your work order management. I have you start understanding who's in what space, when our space is clear, what our space is being utilized and how do you prioritize which issues are resolved based on the severity of the issue who it's impacting, where it's impacting, or what systems is impacting, having that twin can start allowing you to draw context from, pieces of all these different solutions together.

Got

James Dice: [00:31:51] it. I love that. Yeah. And there's a couple, yeah. Different threads that have come up in several of the past couple of podcasts, which is one, is like our state-of-the-art right now, if you just look at the buildings industry as a whole, our state-of-the-art is really, the point solution  or the siloed application.

Right. So you're, you're starting to get to where that is kind of becoming a thing of the past. Right? Um, the other, like aspect of this as the. Easier for an owner to sort of fire a vendor. It is right there.

That's a double-edged sword. So the vendor might not like that, but that same setup allows them to also integrate really fast.

Right. So they can just plug their application directly into this. Existing stack, whereas before they had to do a ton of work and you mentioned people doing the same work, like our industry is plagued by everyone sort of doing the same shit. Right? Uh, everybody's riding the same drivers. Everyone's doing the same.

Same integration. Everyone's, adding their own ontologies off of the same point lists like we're doing so much, that's the same. And so the double-edged sword is like, yes, you can get fired tomorrow. If your SAS product doesn't work. But also you can also plug it into way more, peoples.

Buildings. So I think that's going to be an accelerator in that the best solutions will sort of win out faster. Right. And

Matthew Vogel: [00:33:11] that's an important industry for if we come back. Uh, I remember the first, one of the first questions you're asking me is why is digital transformation, this industry so slow? I think that's a great example of it.

People are used to in this industry, like they know how to run RFP processes. They know how to choose a vendor based on the individual merits of that vendor. You can look at five different building managers and systems or, you know, space utilization systems and you know, and the traditional software checkbox that does have this feature does that this feature do I like the UX and isolation, that makes sense.

And, and the Industry is able to do that. but now it's a very different type of question you have to ask them as part of that process is how does this accrued to my digital twin infrastructure? and what are the outcomes I want to draw? What are the insights I want to take from that solution in conjunction with everything else that I have in my building, the very different way of looking at that problem space, because now you have to manage not just the vendor itself, but also be integration.

Yeah. and so I think that's learnings that we're seeing and that's what people are adapting to right now. but kind of come back to your last statement too. we are seeing a lot more momentum for the people who are jumping on this early, because there is a sort of first mover advantage this from a property owner perspective where, the property owners that are embracing digital twins and this type of integration model are clearly seeing, not just cost savings, but also value adds.

and enough, so that we have partners in the space who are demonstrably saying, I can show that a building with a digital twin is more valuable, an asset than a non digital twin building. so they're seeing it from that side, but at the same time, the vendors who are actually starting to transform their solutions to use these ontologies and to use digital twins, are that much more accepted by these property owners who are mandating it as part of their digital infrastructure.

so we, we do see it taking off very quickly. We see, you know, exponential growth in that area. and it's kind of among the first times we're actually seeing the two ends really meet in the middle to drive the ecosystem and the industry at large.

James Dice: [00:35:13] Got it. You just reminded me of something that I wanted to circle back to from our conversation at real calm.

So. Well, something that came up there was like, how does the building owner gets started? Right. So you mentioned the perspective of a very large building owner and they're coming at it from like, I'm going to go tell all my asset managers that we're going to value buildings differently, but it's something digital twin and like, okay.

That's pretty progressive from, from that standpoint. But like, If I want to get started

tomorrow, the value that makes that asset worth more is I'm implementing all of these use cases, right? I'm touching the occupants, I'm hitting tenants, that's a big undertaking, right.

To go from where we're at today, which is point solutions. At best two, like this fully enabled digital twin experience. Right. So one of the things we talked about and the panel discussion was, do you start with one use case, prove that out and grow it out, or do you start with, you know, just getting all the data into the twin and then it will become clear later on where the value is at.

So I want you to sort of repeat your answer or you can come up with a better one, uh, from what we talked about on the panel,

Matthew Vogel: [00:36:21] Yeah, think it's a huge issue that, people face when they're dealing with digital twins is it's very easy to get lost in the art of the possible, or even just the academia of a digital twin and spend just hours talking about it and the possibilities and all that.

And the possibilities are true and the possibilities are, I guess we could call it endless. That might be cliche, but, the key, to actually unlocking that value is getting started somewhere. and you know, early on in the process for us since we were new to the real estate industry as a company, we engaged a lot with our corporate strategy team to understand how we should interface with the industry and what role we had to play in.

And we actually had a similar discussion, a lot of the world's largest landlords as well. and so we quickly gravitated towards that three 3,300 model that everyone talks about in the industry. So, companies spend $3 per square foot per year on energy, 30 on space, 300 out of the people in the space.

And naturally it's Microsoft and we've seen other companies do the same thing. We jumped on that 300 number. We said, that's the biggest number on the slide. We said, we're the most experienced there because of our background office. let's go tackle that. we quickly realized number one, that's where you get very lost in all the possibilities.

But number two, the value is a little bit squishier there. we know that there is value in saving people time and then making people more comfortable and helping people be more effective. But do you quantify that in salary? Do you quantify that? And you know, time-saver, you know, there's a bunch of variables that aren't necessarily easy to pull up the count that it says, save this much money, or I added this much value.

Whereas with the three number, you could very easily say I'm turning off this light switch, or I am turning off my HVAC, or I'm doing this other thing to save energy. I know exactly how much money I'm saving there. It's a very quantifiable and what we've actually seen on our own campuses. We can actually break even on that within 12 to 18 months.

And so at that point afterwards, you're printing money for your, implementation and you can just keep growing from there. the next piece of that would space is still quantifiable. You know how much, least you have, you know, how much real estate you're owning. it's just a little bit less liquid.

So it's a little harder to impact directly because you might have contracts or you're not going to take a sledgehammer to your wall every day. but we know that there is value in understanding how space is being used and how you can reconfigure it. Especially now, given a lot of the changes with, uh, COVID pandemic.

and so the interesting part of that though, is that it's not a separate solution necessarily. It could build upon the foundation. You've set with your investments in the energy and operation space, because eventually what you're going to change about your space, but also impact the way you use energy.

So again, back to my example of, you can turn off the lights and turn off the heating. If you know that space is typically not used during given hours, they're automating that and start optimizing. Then eventually when you do get to the people in the space, you're also still going to need to know what the spaces itself, the design, the capacity, all of that.

You're still gonna need to know how to interact with the building management system and the controls system in your building. But now that you've added those concrete players in the three and the 30 space, you can more easily tackle the 300. So the, partners and the customers that we've seen be more successful in this space are able to very clearly define what their business goals are.

how they want to achieve that and also lay out a very clear plan about how they're going to go tackle it, because it, I think people should dream about the different possibilities. They have three, five, 10 years down the line, but if you don't get mobilized on something realistic and something concrete in the short term, you're just going to get lost.

In theory for years. so we've seen people basically like RXR, for example, carved out a lab within their headquarters. So they could start implementing there first and start experimenting. But then they had a very crisp vision about how they start rolling out from that one lab to their broader building, to a set of buildings, to their entire portfolio.

that's a very great example of how you achieve success at portfolio scale, by starting with something very crisp and painting a roadmap about how to evolve.

James Dice: [00:40:20] Got it. I think that's a great answer on like letting ROI sort of drive your, Prioritization of your use cases. And then like the ROI really depends. This is what I say in the course, ROI really depends on how fluffy are willing to accept the fluffiness factor of what you're willing to accept.

And so I think that's a great point. so let's transition to sort of, I want to circle back to DTL because. obviously this is not the only, effort to create some sort of ontology happening in the industry. And so one of the things that I'm trying to do is sort of clarify all of these different efforts, and sort of not really compare them to each other, but say where they fit in the context of the other things that are going on.

Right. So I've done several podcasts now around. haystack brick. Um, the previous one, when this, when this show airs, the previous one will have been Google and their, digital buildings ontology. so we talked about that. Um, so want you to kinda talk about where DTL fits and then obviously real estate Corp.

That's sitting on top of DTL Like we talked about earlier, um, where that fits in this. Grand scheme of things of, of similar efforts.

Matthew Vogel: [00:41:30] Sure. Yeah. So DTL TDL, I'd say it's the underlying modeling language. So it's actually not an ontology per se itself. but it's just a scheme by which you could describe any entity as well as the relationships between them.

So it's just, it's basically it's Jason LD format, and it's just meant to say, this is a structured way of you, describing the name, the capabilities, the properties, the telemetry of your entities and the connections between them. that's actually all open source. It's not something proprietary that we want to own.

It's just a way of us saying this is a structured way of, the ability for us to bring data into Azure, digital twins, such that you can model any entity you want. as a digital twins in turn has parsers and capabilities to take DTL and make something of it. But there's nothing stopping people externally from looking at DTL and making their own parsers or toolkits around their own platforms to inherit that as well.

Yeah. then you start layering that up with ontologies or sets of models on top of digital twins, definition, language, which is where real estate core comes in for us. we spent a lot of time trying to figure out do we come up with our own ontology? Do we inherit from a few ontologies? what do we do there?

And. For us, it was extremely important, not to reinvent the wheel both because we didn't want to spend our own time, you know, doing that when there's obviously a lot of learnings in the industry. also because we didn't want to be like the XKCD comic where there's. Yeah, there's one slide. There's 15 competing standards.

I'll build a new standard that addresses all of them. And then fast forward, there is 16 competing centers. Um, we didn't want that to happen. We wanted to leverage what was there already. And, and real estate Corp was something that was provably, adding value to the industry in a way that, was actually used in practice in the real world.

so the reason why we went forward with real estate Corp, not just for the merits of the ontology itself. Was also because when it's used in practice, it's used as is, uh, when we looked at things like brick and haystack, people tend to have their own flavor of it. Your version of brick might be very different from my version of brick and it would be really hard for that property manager to integrate both of those together.

Totally. With real estate Corp, they actually have an existing set of partners and customers who are using the ontology as it. So one of the examples I gave in my I become blog was, I knew which builds up prop tech O S solution, and some of their key customers are people like Basset Cronin, which is one of the biggest, construction and property managers in Europe.

There's Y it, which is one of the biggest facility managers in Finland, and they also operate across Northern Europe. and so they consume that solution and they consume it using the models as they are described in the ontology itself. So we, we no longer have that dialect issue because, it's used kind of to the letter of the models.

Cool.

James Dice: [00:44:17] And, and real estate course something that's not as well-known in the U S it doesn't come up a whole lot. Right. So, it might make sense to sort of give a little bit more background on how that relates to, like you mentioned, it's used as is, but how is it able to be used as is, and is it, an open source project that's happening just elsewhere that people can contribute as well?

Matthew Vogel: [00:44:39] Exactly. Yeah. So it's open source. They have a kind of consortium that looks at the contributions and make sure that, the models kind of adhere to the overall connections and the overall, Voice and kind of tone of all the different models and how they fit together. and then what they did very purposely from the start was to make sure that they had developers and customers lined up to use those models directly.

So they have, uh, a really interesting blend of academic stakeholders, technology stakeholders, and real estate stakeholders, all of whom can validate, the ontology end to end. Um, they, they came to us because they're interested in digital sprints a long time ago. but we were always intrigued by the idea of an industry leading ontology.

And so as we held the capabilities to, open up DTL to allow any modeling, that's where it became much more possible for us to go back to them and say, Let's actually build the next version of real estate Corp using DTL. Uh, and so now you can actually use the DTL language, to leverage real estate core, but then your actual digital twin solution.

So there's nothing stopping people from still using real estate Corp, as it was previously built in their own solutions, but now we make it that much easier to connect it into anything at your IOT based because we'll know how to kind of adhere to that TTL contract.

James Dice: [00:45:56] Got it. I love it. So when I

read your blog, by the way, and we'll put your blog in the show notes, I want to circle a one sentence in it, which said.

Real estate Corp specifically does not aim to be a new standard, but rather provides a common denominator of bridges, other standards, such as brick project haystack and more. So is it like translating between

different ontologies or how does

that, what does that mean?

Matthew Vogel: [00:46:18] Yeah. So, when they first started out, they actually do take a lot yeah.

Concepts from things like break in haystacks that way an owl, as well as another language that they can bring in. So they do have an ability to translate models externally into what they use today. and so that was also an interesting concept that they, you could have a bridge between, ontologies to help with your on ramp and accelerate your time to results again.

but I think where it's also, more interesting for us is that, once you're in that ontology, you could use it directly in the tooling that you have. and you know, in our version of the world, its DTL and our partners and the national digital twins. and so again, it was kind of a one-off translation.

We did to make sure that real estate court could be built with DTL. But now what we've proven is that you no longer have to, translate or convert. You could just kind of start with that tooling as it exists in the repo today. at the same time, we do recognize that there might be blind spots that we have.

There might be gaps. There might be areas where we need people to bring in their industry expertise. and so what we worked on with the real estate core group was ensuring the plugability of the ontologies that way. Let's say if we don't have elevators, so we don't have deep expertise on elevators, you know, apartments like disapprove elevator or, others might be better suited to describe the entire model of an elevator in the systems and data could send you, um, that's where we allow them to contribute their models to extend, but make sure that it's not Island.

It's not, they're not a siloed ontology. It fits together within the right asset classes or equipment or within the broader building structure.

James Dice: [00:47:53] Got it makes total sense. So how are you seeing, I know you guys have a lot of projects going on right now, or how are you seeing a building owner say, I want to standardize on this

ontology.

Like how do I, as a building owner, make sure that this happens in my building.

Matthew Vogel: [00:48:09] Yeah. So, I think, again, it starts with the business value. You're trying to drive because even though the intelligi is very complete, it's pretty rare for anyone to use an entire ontology, in any given solution.

You're going to take a subset of that ontology for the. Pieces of the domain. That makes sense for you. So the parts of the ontology, you use her as space management solution. It's very different than the ones that you'll use for an operations, solution. so the first thing is making sure you have your clear view of the business goals, but then you can start kind of driving conversations with partners and customers about.

what are the solutions you can provide and how do you plug into my infrastructure? So a lot of the bigger property managers who've seen. So if we come back to the logging in, I mentioned Brookfield and Oxford, they basically say they want to drive the holistic digital twin infrastructure of their buildings.

First that way, they have the right backbone to kind of hang these other solutions onto. and they basically say that in that solution, we want, our ontology, our domain expertise to be real estate core ontology, because moving forward as they engage with specific partners or specific solution providers in the industry, they can say that here is the structure I have in my building.

Here's how I will allow you to integrate with it. how can you go plug in with me? Uh, in some cases, we find that the opportunity is big enough for a solution provider to completely rearchitect their solution, to use the ontology natively in their software as a service model. In other cases, we've seen, vendors, whether it just as a kind of quick and dirty implementation or desperate time or resources sake.

provide a translator on eat grass. So that way, even if the entire city, the solution isn't rearchitected, you can still translate into the data that your consumer expects the data format to be in.

James Dice: [00:49:55] Got it, this whole, platform aspect and the value of it. I've struggled to integrate it into my course, because the course is introductory.

And I want to sort of paint the picture of like where the industry is at today, but it's almost like I need a separate module. That's like, we're used to integrating and creating these point solutions that are sort of full stack. there are building owners out there that are just going the platform route and saying, okay, now

your solution will

integrate back to me.

Uh, and I haven't quite figured out how to, deploy that in the course. So yeah. That's advice. Anybody has advice. Let me know,

Matthew Vogel: [00:50:29] uh, as we. Good analogy I've also seen happen is like take PCs today. We'll take for granted the fact that you could plug in a keyboard or plugging a mouse or plugging a monitor, and it just works under the covers.

That's actually happening is some sort of plug and play type logic where your operating system has dictated. This is the contract by which I will recognize your equipment. If you say that I am a monitor, I am a keyboard or I am a mouse. And these are my capabilities. I have a scroll wheel. I have to click buttons.

I have assigned click bullet. I don't know exactly how to pull in your drivers. I don't know exactly how to command it. Control your device. Now you take that same concept into a building and you have the building owner saying, this is the operating system in my building. it could be a Willow, it could be a Bentley, it could be magic.

You know, any of those players who are targeting the holistic twin of a building, then you could say, that's my contract. That's my integration layer. Here's how you can go play in this ecosystem. And so now you can start driving more of those solution providers to better plug and play with your building.

James Dice: [00:51:32] Totally. I love that. Cool. So as we kind of close things out, I wanted to ask you and, I put you on the spot here. What are you sort of excited about? I know you guys just made several announcements, but if you look a couple of years out, uh, what are you excited about on the Microsoft product side of things?

Matthew Vogel: [00:51:49] Um, I think one key thing is making sure we keep accelerating time to results. So I think now We've shown that there's value here in that ontology and in the ability to plug and play these solutions into the digital twin. the magic has to be in shrinking the time to results from again, months into days or hours.

We need that to be as quick and seamless as possible for any of this to drive, at scale value. we've seen now it's, not easy, but at least there is line of sight and straightforward for anyone in the industry to go tackle one building at a time with spot solutions. Yeah. Uh, for someone to succeed, it's going to take thousands of buildings and showing that you can do this at portfolio scale and minutes or hours or maybe days.

that's one thing that I'm really excited about. And then the other piece that's exciting for me is. from a Microsoft perspective, we're seeing all of these worlds, from our, industry collide together. So Microsoft is unique in that it has that three cloud story. It has, Azure, it has an Azure IOT.

It has dynamics and it has office. This industry is perfect for that intersection. And so I do see. a lot more of that value coming to property owners, tenants, uh, workplace solution providers, and so on, to leverage all of the capabilities from Microsoft, for any given solution to the industry.

James Dice: [00:53:09] Well, cool. I appreciate you coming on the show. I think this'll be a great follow on to Emmanuel's episode, but also a lot of the other ones as well. So, thanks so much and we'll talk to you soon.

Matthew Vogel: [00:53:20] Yeah. Thanks for having me.

James Dice: [00:53:21] Alright, friends. Thanks for listening to this episode of the Nexus podcast. For more episodes like this and to get the weekly Nexus newsletter, please subscribe at nexus.substack.com. You can find the show notes for this conversation there as well. As always, please reach out on LinkedIn with any thoughts on this episode.

I'd love to hear from you. Have a great day

⭐️ Pro Article

This article is for Nexus Pro members only

Upgrade to Nexus Pro
⭐️ Pro Article

This article is for Nexus Pro members only

Upgrade to Nexus Pro

Are you a Nexus Pro member yet? Join now to get access to our community of 600+ members.

Join Today

Have you taken our Smart Building Strategist Course yet? Sign up to get access to our courses platform.

Enroll Now

Get the renowned Nexus Newsletter

Access the Nexus Community

Head over to Nexus Connect and see what’s new in the community. Don’t forget to check out the latest member-only events.

Go to Nexus Connect

Upgrade to Nexus Pro

Join Nexus Pro and get full access including invite-only member gatherings, access to the community chatroom Nexus Connect, networking opportunities, and deep dive essays.

Sign Up