Remember Google Earth? It still exists today, of course, but remember its heyday in the late 2000s? It was such a wonder to zoom in from its home-screen globe down into different parts of world, to fly over whole cities rendered in 3-D, even if the graphics were basic and the details largely blurred.

Yes, it was a lot of fun for armchair travelers, but it wasn’t of much use to the AEC industry, and Michael Jansen knew there could be something better. The architect-turned-CEO of major BIM-services firm Satellier acquired an early 5-D BIM company, Screampoint, in 2008 and hired one of the creators of Google Earth a year later to better harness the wide-ranging capabilities of the technology. Now Jansen’s chairman and CEO of Cityzenith, whose 5D Smart World software suite is officially set to release commercially on May 3 with far more detailed and adaptable maps of one hundred of the world’s biggest cities—with more cities and regions soon to follow. Users can zoom in on detailed 3-D renderings of individual buildings and streets, analyze them through terabytes upon terabytes of open-source and real-time structural and environmental data, and import BIM models of their own and manipulate city grids and structures to see how their projects might fit in and interact with their surroundings.

“A flurry of innovations in tech in the last two years have made this possible,” Jansen says. “The timing is fantastic. We often walk into offices, and we have the principal sitting across the table, saying, ‘We’ve been waiting for someone to walk in with this for the last ten years.’”

The product’s launch will take place in association with the BuiltWorlds Summit at the UI Labs Innovation Center, and as a preview, Jansen recently sat with us to discuss how he came up with the idea, what all Cityzenith’s software is capable of, the wealth of data it contains, and what it will take to expand on that data.

So, you began your career in architecture. How’d you end up transitioning into software?

Almost by accident. I studied architecture at Yale and then Cambridge. In the early ’90s, right after graduation, I flew out to China and started with a major US architectural practice as a twentysomething junior project architect, being thrown on to massive Chinese urban-core regeneration projects that were twenty years too sophisticated for me. It exposed me to the inner workings of the city.

Around 1999 or 2000, I had moved from China to India, which at that time was thought of as being the next China. I worked there for another major US practice until I started Satellier, a CAD-outsourcing business. That company started with me and four architects in the third bedroom of my apartment, in New Delhi. Four or five years later, we were 500 hundred people spread across five floors in two buildings, with Sequoia Capital as investors, and we were serving 30 of the top 50 AEC firms in the world. Satellier became the largest BIM-services company in the world at the time.

Is that when your idea for a comprehensive structural mapping program first took root

Because we were making BIM models all day and all night, we became familiar with the power of 3-D object modeling, the power of this semantic 3-D tool to better organize, coordinate, visualize, and analyze data. Around 2007, we made a decision to acquire a company called Screampoint that had the beginnings of this type of technology, something they called 5D. We bought the company in 2008. I became chairman then and later took over the company as part-time CEO in 2010 and then full-time CEO in 2012.

The first order I gave as chairman was to hire as CTO the creator of Google Earth, Rémi [Arnaud], who came in between 2009 and 2012 to prototype the first iteration of something that we then called 5D Smart City. We learned a lot through that process. It was about discovering the business model, the right technology approach, all the fun stuff. The idea that the world would need Smart World was self-evident; we just weren’t sure what the route to market was. By 2013, we began talking to cities, to really try to make this core platform technology interoperate with everything that the city had. We launched with ten major cities around the world, from Singapore to Barcelona to Chicago, and worked with those cities for three years to effectively build out the platform.

Your videos show the basic capabilities of Cityzenith’s software, but could you describe the full scope of its capabilities?

All along, the vision was that this would become a universal platform for the building industry, a kind of universal data-exchange and data-analysis platform. For our users, the moment they log in, for their $99 a month, they get over a hundred fully loaded 5-D Smart City models in 12 countries. They get thousands of curated public and commercial data sets, and they get dozens of free and premium analytics tools.

Users take their 3-D project models, as they’re working, and they’ll drag and drop them into Smart World. They take that 40-story SketchUp model, drop it on-site, position it using the coordinates, and then begin to run analysis. They do everything from investigating Energy Star reports of neighboring buildings to looking at bike routes to studying the impact of traffic and transit on building design. The look at shadow studies and line-of-site analysis, and they get more complicated tools like solar-rooftop analysis and radiance studies from glass curtain walls. They can actually study the refractivity of their buildings on neighboring buildings.

“You have master planners doing a $400 million project using census data from ten years ago to predict people movement. It’s obscene. They should be looking at our sensory data showing exactly where people are moving as we speak.”

What are some of the benefits for users?

Users that have worked with the tool have reported savings and a reduction in man hours for time allocated to project research, team communication, and design and team-planning exercises. They’ve also eliminated hundreds of thousands of dollars in annual spend on subscriptions for legacy software tools.

The analysis function allows you to take data—both your own as well as all contextual data—and run any type of analysis related to your building project. The collaboration component—because this is also web-based—allows you to work with your entire project team at the same time, on the same project site, on the same model. They can run analysis right in front of their clients online. The client can say, “Add add five stories to that building,” and the architect will click on the planning guideline tab and pop up the zoning-envelope feature and reveal all the zoning envelopes for the city in the 5-D model and say, “I’m sorry, sir, but if you do that, it’s going to stick out by four extra floors.”

John Kizior, from AECOM,  wrote a beautiful article a few years ago called “Information-Led Design.” That’s the future of AEC for me. Right now, architects are designing in a vacuum. You have master planners doing a $400 million project using census data from ten years ago to predict people movement. It’s obscene. They should be looking at our sensory data showing exactly where people are moving as we speak.

And in addition to designers, what are the advantages you foresee for contractors and owners?

Contractors have been talking to us about construction-logistics management, especially for large projects, because when you’re doing a big project, you’re pulling in materials from all over the place and managing dozens and hundreds of parties and increasingly leveraging geospatial technologies to optimize construction logistics. Vermeer and Kiewit were companies that took part in the beta that did just that. For owners, we’re pulling together all their real estate asset data and combining their BIM models along with everything else they have so that they have a single global portfolio-wide asset-management tool.

The play is the same for everybody: consolidate, curate and normalize, visualize, analyze, predict. Where we are as a company is step four; where we’re headed is step five, which is predictive outcomes. If I move the entrance to this building from this part of the site to this part, what is the impact on traffic? Given the bike routes around the building, where’s the best place to put an entrance? What is my building’s influence on the surrounding microclimate after construction? Two years from now? Five years from now? All of this we can model with predictive tools that we’re now building.

Talk a little bit more about all the different sources you’re pulling data from and how they interact with each other.

Much of it is publicly available open data. That can include everything from satellite imagery to 3-D building models to infrastructure data to Energy Star reports to building permits to various types of web services such as road networks, traffic, weather. There’s also a growing stream of premium IoT data from partners that we have like GE, Huawei, Microsoft, and others that have sensors installed. Our city models become more robust literally every week.

And what about the underground systems shown in some of your videos?

The underground 3-D model assets do not feature in our models. Those are extremely secure projects that we participate in, but at the moment, that’s not available through us as a service.  Eventually, the goal is to be able to have our client firms be able to leverage our models for their underground capability as well. The technology can already do it, but in terms of actually getting that data, right now there are only two blocks in all of Chicago that have been modeled. So, it’ll be a while before that comes. The project there is the [UIM (underground infrastructure mapping) pilot], by City Digital and UI Labs, and it was recently unveiled by Chicago CIO Brenna Berman to Chicago Mayor Rahm Emmanuel and London Mayor Sadiq Khan.

As this software opens up, will you be relying on the participation of users for any additional data?

We’re a bit like Dropbox, so whatever, say, an architecture firm does with our model, we’ll never know. A lot of architecture firms have their own project data, but they don’t want their BIM models shown to everybody, especially while being designed.

Having said that, the moment [a design] gets approved by the building department, we have a feature where every single building that is not only up but is coming up is also going to show up as a layer. So, if you’re designing a forty-story tower, you might want to look at that because you might find out that a few years from now there’s another tower coming up right across the street.

“We can work with any data you’ve got: any API, any flat-file data, any 3-D model.”

Given that you’ve worked with one of the main creators of Google Earth, is the idea to eventually have a Cityzenith map of the entire world? 

More or less. As far as we can go until somebody fires me or I die.

About how much do you have mapped out with data already?

We’ve modeled over a 100 cities so far. Some are much further along than others, depending upon availability of data.

Is it more difficult to conduct data collection in certain regions or cities?

Absolutely. It’s a lot harder in countries that don’t have robust open-data programs. And how we get around that is, again, in theory, you as a user can populate our models with all of your own data and not need anything from the government. The idea is that we’ll give you everything we know available, and if you have a request for some unique data set, we’ll try to find it. But, you can also populate it on your own. You could, in fact, build your own Chicago if you wanted to. The tool works like a video game. You can knock down a building and put your own in there. It’s completely manipulable.

But we went to market with one hundred [cities] because we wanted to say, “Hey, we’re it, and there’s not going to be another one.” In that one hundred cities, we cover the world’s top AEC sectors—London, New York, Chicago, every major one. I think, in the next three years, we’re going to do most of our business in two or three hundred cities. And we’re constantly adding cities, constantly adding data.

How do you plan to start collecting data from midtier cities and smaller markets?

You work with what’s available. It’s worth noting that there are over thirty countries—not cities, countries—with robust open-data programs in the world today, and that trend continues. In terms of smaller markets that may have less data, I think we just haven’t come across that yet because we’re still in the big markets. Chicago’s one of our models, and Chicago’s a big city, but it’s number 83 in the world, in terms of size. We have a long way to go with what we’ve got.

And are there other types of data that you hope to incorporate that you haven’t yet?

We definitely want as much real-time data as we can conceivably get our hands on. The more dynamic the data, the better the predictions. We already have thousands of existing data sets, and as IoT and sensory installations become more popular, we definitely want to make a beeline for that stuff. Our Chicago model will soon have all of the AoT [Array of Things] data from the University of Chicago’s sensor project, which now has about forty locations with four active poles.

Because of your prior experience in BIM, I’m sure you know how frustrating it can be when different software solutions in the AEC industry don’t talk to each other well. Are you hoping to engineer Cityzenith to connect directly with any other software solutions?

There’s a tool we’ve built called InstaBIM, a universal BIM modeling converter. Whether you have an FBX or an OBJ format, IFC, SketchUp, Rhino, any of these 3-D BIM formats out there, you can drag and drop them into Smart World flawlessly. And you can choose what data to bring in with those different models.

There’s another tool called Warp, on the contextual-data side. In theory, we can import any file format— geoJSON, Excel, Word, PDF, all of them—into Warp, and in Warp you can normalize that data set for Smart World by assigning categories to the data fields. You can change things around, you can index it, and you can tag it so that you can search it better later. We can work with any data you’ve got: any API, any flat-file data, any 3-D model.