Jack Dangermond is the founder and CEO of ESRI. ESRI’s software is used by every level of government around the world. You can see ESRI’s influence in online mapping tools from Microsoft, Google, Yahoo! and FortiusOne. I had the opportunity to interview him over the phone on April 20, 2009. In this portion of the interview we discuss the history of GIS and online mapping.
Jack will be speaking at Where 2.0 on May 20th in San Jose. You can use whr09rdr for 20% off at registration.
Brady Forrest: So ESRI is known for creating large enterprise GIS applications. And then a couple of years ago, maps came into the hand of just about anybody through Google Maps. One question — I asked people for questions for you, and I think my favorite one was: Do you think that the explosion of web-based mapping has just filled the world with ugly and poorly designed maps?
Jack Dangermond: That’s an interesting and compelling question. I think what the consumer mapping sort of technologies have done is provided geo-awareness to everybody. And they’ve done it principally by building a standardized basemap for the planet. And Google has been, obviously, the leader in this. But also, Microsoft is making a lot of contributions in the same space. And that allows me to navigate and understand a kind of electronic map or image so I can sort of see things. And that’s pretty powerful actually for just spatial awareness and georeferencing people’s minds about what’s going on at different locations. In the last couple of years, I’ve seen a lot of VGI or volunteer geographic information added to it, little dots on maps. And that allows people to participate in these websites and share their knowledge in some kind of a collaborative space. And that’s emerging quite nicely.
But my field is really in geographic information systems. And while we do have a lot of large users, like you said, the history on it has been a little different. It started basically as a university and, I guess, frontier area where people were playing around with computerized mapping for doing spatial analysis. And it started really on mainframes and then went to minicomputers and workstations. And where it really found its way was on the PC. So my judgment is that there’s at least a few million users that are in the business of creating maps or creating spatial analysis and then using them for real applications like sighting or like environmental analysis or healthcare studies or land use planning, that sort of thing. The explosion that happened on the web was sort of with the invention of putting maps on the web for many people to look at. And the big popular ones were way back in the time of — I say way back, but it’s in the time of MapQuest when they became really popular, that notion of navigating and seeing. The actual basemaps of Google and Microsoft are less about creating new authored maps than they are about visualizing stuff on top of the basemap. GIS actually has a little bit of an interesting culture because there’s lots of people that make really crazy maps in GIS on their desktop or now increasingly on the web. They mix colors wrong. They do color palettes incorrectly. They mix symbologies incorrectly. And in the mashup, it’s just exemplified. If that’s what you’re thinking about the web is just exemplified people making crazy stories and mixing relationships, yeah, probably right. There’s bad cartography out there.
What I’m interested in is that a few people who are really talented and special can build templates for cartography that other people can use. I like to think of it as sharing geographic knowledge. So in the early years, people shared geographic data. You know, my layer and your layer and they could combine them and so on. What I’m seeing now is the ability to share a cartographic template. Like here’s a wonderful color ramp that you use for demographic data. Or here’s a standardized template for how you display cartography to create a basemap or a utility template or, you know, like that.
Brady Forrest: Who are some people doing that work?
Jack Dangermond: A lot of my colleagues at ESRI are doing it. They’re actually designing ways to build templates and share them, that is encapsulate them and then share them on the web. One of the first ones we did was more for engineering applications. It’s water hydrologic networks or water networks like the water utility would have. But the clues are there for how we would leverage that for all cartography: basemap cartography, vegetation cartography, geologic cartography and so on. And it’s very similar in concept to the idea of standardized data models. In the last ten years, we saw a movement within the geo field to move from everybody does their own data model to having standardized or industry-wide cartographic model — database models that say, “Okay. Here’s the ontology for geology.” Or, “Here’s the ontology for certain vegetation or land use.” But this moves it up a notch to say, “Here are some standardized ways — template ways to represent this data with mapping styles.”
And if you looked at the last release of PowerPoint from Microsoft, it has built into it templates. And I like that very much because a few people thought through how to communicate graphics very well. They built it into the product that shipped in volume. And now you see all kinds of interesting but also standardized PowerPoint templates that people use and they don’t have to spend all the time, nor make all of the terrible graphic mistakes that people made with PowerPoints over the years. So the way I see it is that our field will evolve to not only sharing data on the web, but also sharing maps on the web. There’s at least 50,000 different mapping websites now, and there’s probably many more than that. But also sharing the templates for cartography behind the maps. And by the way, as long as we’re on it, also sharing analytic models that one user can create. Let’s say a famous scientist creates a model. He shares it on the web. Other people can download that analytic model and reuse it by combining different map layers in such a way that will make a suitability map or a flood hazard map or a where to grow corn map or a hurricane prediction map. And those kinds of maps are not for — I mean the models that are associated with the maps are not so simple to think out. In fact, often authoritative source people are their authors, then you want to use those authoritative sources because they know something about let’s say water resources or geologic hazards.
So if one person encapsulates the model and shares it, other people can use it and apply it with their own data and their own location. The same philosophy as the cartographic templates. You get some data here or some data there. You associate it with a standardized way to display this data. And you can turn a layperson into a rather sophisticated cartographer.
Brady Forrest: Can you dive into a little more about the analytic models? What would be the simplest example of that? And is there anywhere on the web right now where people could take advantage of that?
Jack Dangermond: Yeah, there is. There’s lots of it. In our software, our GIS, there’s a capability to develop models. It’s called a model builder. And literally tens of thousands of people build models. An example of a model is one done by David Maidment. David Maidment is one of the greatest water resource analysts in the world. He’s a hydrologist. He’s at the University of Texas in Austin. And he’s taken all of the layers that would be necessary to describe surface water flow, now he’s working on groundwater, but his basic models are called arc hydro. So he takes rainfall. He takes slope and impervious surfaces like land use and vegetation; calculates runoff. Takes things like the elevation and hydrological network and he flows water through the network. And he calculates flood plains. And he does that with a GIS.
So he’s published that in a book called Arc Hydro. But he also has his models available on the web. So you can Google David’s name and find these models; download the models. Associate them with your layers of soils or vegetation and so on. And run a very sophisticated hydrological model but on your geography. So I like to call it hydrology — it’s like Turbo Tax for hydrological modeling. Many people use Turbo Tax, and they trust that people who are authoritative source people have thought through how to do taxes. And so you type in various perimeters, your tax and income, and out comes your forms. Same idea here. Because of David’s experience and his documentation, the model, many people will trust David’s model and apply it. USGS does this. The EPA does it. Many nations around the world do it as well as state and local governments and engineering companies. So this is a way to share a kind of knowledge set that is a combination of various geo perimeters and have it come out with standardized expression.
Brady Forrest: What are the social uses of this? Is it used with like say electioneering?
Jack Dangermond: David’s model?
Brady Forrest: Not David’s but someone else’s, like how they’re —
Jack Dangermond: Oh, people do it for forest management. They’ll predict the complex forest change, growth, succession and so on. So they’ll build a vegetation succession model. People will do it for population growth and change or change in social neighborhoods by looking at changes over time. The oldest of these were simple population forecasts and people like Wendell Bell and Chomsky modeled and projected change in social neighborhoods. This was way back in the 50s and 60s. And applying that within GIS gives you the ability to model change or landscape processes. Not simply look at the map, but also model behavior of physical and social landscapes.
I mean people actually in police and law enforcement do this for modeling behavior of crime or it’s called crime analysis. They’ll look at different perimeters of incidents of crime. Look at it in terms of a social networking or community environment and then forecast possible crimes. And then as a result of that, police departments will redistrict their law enforcement resources to most maximize where these issues are.
This was the first portion of a three-part interview with Jack. The second and third are on Sharing GIS Data and the Geoweb, respectively and will be out this week.
The first question was courtesy of Chris Spurgeon.