As the president and founder of Esri, the world’s technical and market-leading producer of geographic information systems (GIS) software, Jack Dangermond has guided that company’s growth from its founding as a research group in 1969. The recipient of no less than three collegiate degrees with emphases in landscape architecture (B.S. Cal State Poly, M. Arch. University of Minnesota, M.S. Harvard), Dangermond became heavily involved with GIS as a landscape architectural graduate student at Harvard in the 1960's.
After its early days as a research group focused on land planning, Esri began to focus more and more on the development of GIS software, eventually releasing its first commercially available software program, Arc/INFO in 1982. In 1999, Esri introduced a version of Arc for Windows operating systems that has since become commonplace in workplaces worldwide, influencing fields as diverse as geology, fire and safety, police enforcement, architecture and engineering, city planning, and many more.
Can you briefly tell me a little about the early days of GIS, who was advocating it, who was using it in the field and how that's evolved over time?
Yes. When it started, it had several birthplaces. One of them was at Harvard at their lab for computer graphics and spatial analysis. That was in the mid-60s. A professor there by the name of Howard Fisher founded this lab for computer graphics. And one of his early researchers was Carl Steinitz, who was a Ph.D. from MIT who got interested in computing and computer graphics and then took on the development of software graphics systems called synapse and GRID, and GRID was designed to do raster geoprocessing work and raster mapping.
Carl's interest, initially, was in urban planning but he became a professor of landscape architecture and taught there for about 40 years. The lab for computer graphics pioneered computer mapping, initially with printer maps and later plotter maps and so on. But it also pioneered geographic analysis or spatial analysis, where you would do map overlays to do interpretive mapping. Carl's work largely involved urban simulation, urban plans, visual planning, and landscape planning with computers.
Was Carl consciously incorporating [landscape architect] Ian McHarg's work, was there a connection between those two or were they completely separate?
Carl certainly knew of Ian. And Ian would come to Harvard occasionally and give talks. If you take a step back, there were several big landscape planners. One of which was Phil Lewis who had an approach called corridor mapping, an approach that was out of Wisconsin and then Ian was out of Penn, he did the plastic overlay, thematic mapping approach. There was a guy down in Australia called Christian - two of them actually, Stewart and Christian - they did terrain mapping and an integrated approach for resource planning. There was a guy up in Canada named Angus Hill.
Those guys sort of preceded computers. They did landscape analysis, landscape planning. Carl came at this from an urban planning and computer mapping perspective and certainly used Ian's approach of map overlay as one of the application areas.
The second big home for GIS's origin was Roger Tomlinson in Canada. Roger created the first geographic information system in the world. It was called the Canadian Geographic Information System or CGIS... It was not focused on design. It was focused on inventory and interpretation of geography and was designed to put into a computer all the resource inventories for all of Canada. Things like soils, water, forests, ag lands and then to do interpretations of those for regional planning.
The third origin wasn't really GIS. It was computerized mapping. That was David Dickmore over in the UK. And then there were probably a dozen other ones, people who did what I like to call computational geography.
They were experimenting with the use of computers and maps and models and analytics. They started a kind of revolution in geography, sometimes referred to as quantitative geography. They invented projection systems with mathematical tools. They developed computerized ways to look at spatial interaction between populations. Bill Warns was an example of that.
Another famous theoretician was Waldo Tobler, who is still alive, a very famous quantitative geographer. Duane Marble invented some of the first transportation models. Britt Harris, was also a transportation modeler.
All these guys were working in pockets of isolation, once in awhile publishing a paper. But they all contributed and laid down a lot of the theory and foundation for what ultimately became geographic information systems.
In the 60's I was at Harvard, I was influenced heavily by Carl, his whole quantitative approach, and by Howard Fisher. I moved back here [Redlands, Cal. - Ed.] in the late 60s and started ESRI which for the first ten years was focused entirely on GIS projects. We didn't call them that at the time, we were assisting some landscape architectural firms like EDAW and citing some transmission lines or in other cases we were doing project work.
So you were operating as consultants and not software developers per se?
Yes but we had several programmers. We wrote some of our own software, initially, for our project work. And very gradually we built that software up. But it's what I would call today homemade software, primarily focused on doing projects. And these projects got larger.
We started with things like locating ski runs or locating a transmission line corridor or locating a new town or doing a coastal zone plan. We ourselves weren't doing the planning work but we were doing all the mapping work for the landscape architects and planners who would subsequently incorporate the maps into their actual designs.
After about four or five years, agencies began to get intrigued by the idea of digitizing the maps and then using this GIS system that we would develop on an ongoing basis to have living plans, so that they would continue to use and reuse the same information for different kinds of applications.
At that time there were three basic technologies, there were image processing tools, for capturing and processing remote sensing data, there was CAD, computerized mapping, computerized graphic technologies based on design software, then there was GIS software, which we were one of many who were trying to develop tools that would work in that space. After about ten years, many of these customers, some of them were whole states or whole countries, asked us, could you please give us your software. So we had to disseminate our software as open source mode.
And when was that?
That was in the late 70s, and the result was not very good because you'd have to have a
programmer to run these tools. So after a few years users asked if we could have a user meeting, which we did in 1980. We had 11 attendees from different countries and different agencies. And at that time they were quite interested in us becoming a software provider instead of just consulting.
So we shifted our philosophy from being a computer mapping group that would support planners to the idea of building actual software that would be well engineered. Because at that time our software was not well-engineered at all, it was basically built with project funding and for project work, largely by ourselves. We decided to do it.
Very gradually we re-engineered our software thanks to some really brilliant software engineers here. It took about a year and a half and we came out with our first version of ArcInfo. ArcInfo didn't do very much in the beginning, but what it did, it did it very well. Further development was driven over the years by the user meetings which has grown into our yearly users conference attended by more than 14,000 users last year. Each year our users come together and give us feedback on what we do and we tell them what we're doing to add new features and functions to the tools based on their direction.
We have gone through many cycles here starting on minicomputers and then we moved over to workstations and then into PC versions of our software. And now the whole thing is morphing over to the web.
This past summer, we released ArcGIS online, which has about 100,000 maps from all over the world. [It’s] kind of what you might call Facebook for mapping. People make their own GIS maps and upload them and then other people can search and discover and download those maps or use them in a web environment. Also, there are many web services like what you'd see in some of the consumer mapping and visualization tools from Microsoft or Google. So the Web is our fourth software platform.
So did you see the real growth occurring when you switched from workstations to PC?
Yes, with each step. When we were focused on minicomputers we had probably about 600 to 700 customers. When we went to workstations, we went to thousands of customers, both individual users and organizational customers. When we went to PCs it again went up orders of magnitude. Today we have about 300,000 organizations who use our software, maybe a million seats in the desktop space.
And what we're seeing as we move to the web is a similar phenomenon. But instead of one order of magnitude it's going to be multiple orders. Today, people are using ArcGIS online to make millions of maps a day. Some maps are quite simple, just a topo map or some dots on maps, but people are also creating mashups using GIS. And they're really easy to do, if you have some time and you're technically literate, just go to ArcGIS online and you can find maps and play around with them.
And what fields, or what professions, have you found use GIS the most?
What I want to say is that the field has been dominated largely by geographers and planners, foresters, water engineers, policemen, crime analysts. In other words, specialists. There are at least twenty-five different specialty user types. A couple of years ago, my colleagues and I began to realize that we had drifted away from design. So we started to work on something we call geodesign.
Landscape architecture is basically geodesign, it's designing geography. And yet geodesign is not only done by landscape architects, it's done by some of the world’s largest corporations. For example, Starbucks is creating geographic strategies for where they locate or close stores. It’s also being done by Walgreens and PETCO. These are all our customers. And they're… getting the geographic advantage by using GIS to look at factors and model these factors, make maps, and then they design strategies based on top of them. Foresters do the same thing and so does the military, they lay out military plans of action, strategic plans using GIS. Oil companies use geodesign, and geoengineering to make decisions about where to search for oil.
So GIS in the first years was dominated primarily by people who wanted to do inventories and interpretive mapping, like suitability mapping in a “McHargian” way. But the last couple of years we've been trying to pioneer sketching tools that allow people to sketch on top of GIS maps and very quickly analyze the implication or the impacts of different sketched alternatives. This can allow a designer to incorporate all the science information of the GIS and very quickly look at alternative plans and understand their implications or the consequences of these plans.
Landscape architecture is all about that right now. But it's a new concept for many fields, fields that don't really have the design discipline integrated into them. So I have high hopes that GIS will become increasingly relevant for landscape architects as we make the tools easier to use for the design process of just inventory and mapping. But I also have great hopes that all fields that do work with geography will be impacted by this geographic approach of incorporating science into the way we evaluate and ultimately make decisions about our future.
So it sounds like you're trying to move GIS toward enabling a collaborative process where various users can access different functions toward an end product?
Well, like I said, a lot of GIS has been about making beautiful maps and interpreting maps and doing suitability maps like you'd do with overlays. Then they'd get out the tracing paper and lay it on top of these maps to do the design. What we're doing is building the sketching, design tools right into the GIS software. So that literally as you sketch, you're doing it digitally and you're able to overlay digitally these sketches on the other layers of information in the GIS.
Like I'm going to put some houses here - what's the runoff? I'm going to put change the land use here, what's the increase in traffic, what's the loss of biodiversity, what's the increase in jobs, what's the accessibility of the population to commercial institutions, what's the market potential of this location? All of those are, design/evaluation iterations.
Basically what we're building into the base tool is the ability to do that. And we can do that now on the desktop pretty easily. What we're doing is working on doing it on the web so that you can do same-client sketching on top of maps and quickly evaluate the consequences.
So in general currently what do you see as the most important or frequent tasks or functions performed by practicing landscape architects using GIS that couldn't be performed with other applications?
Well I think the small firms rarely use GIS. They use CAD to do their drafting work or
presentation work. But most small firms actually just don't have enough business to do GIS. Specialty firms do. Some have chosen to say "I'm going to use GIS technology to distinguish my practice.” And some of those are doing quite well. There are a number out here on the west coast that do [use GIS], there are some up in Canada, there's EDAW.
And then there's the mid to large-size firms which actually have a GIS practice, and they've done very well by taking landscape approaches to doing regional planning work, and the engagements there are with agencies like forestry or parks or the Army Corps who actually know GIS and they want a more quantitative approach.
People typically use GIS when they have a lot of complexity in the landscape they're studying, when they need to look at a lot of alternatives, where they have to have quantitative information that describes their design. Where they have citizen scrutiny or other agencies' scrutiny or they really have to provide evaluations of the quantitative impacts of their results. Or large areas where it's just impossible to do it any other way but by computer.
So for students just graduating from a landscape architecture program, how essential or how valuable do you think it is for them to have GIS skills, or at least an awareness of how it works?
I think if they're going into public service it's very valuable. Like in government agencies or the forest service or park service or regional planning agencies. Because those organizations basically are doing everything using GIS. In private practice, if you're going to a large firm, my assertion would be that it would be a very valuable skill, it would be a distinguishing skill.
If you're going into a small practice of a dozen people you'll find sometimes that skill being practiced heavily and other times [the attitude is] “Oh, that's kind of a neat idea, but you know, the customer's really looking for the landscape plan, the planting plan, a more traditional approach,” I'm not suggesting that it's not great work, it's just different work than the GIS practice.
And as far as . . . the education aspect of it . . . from your perspective is it being, is GIS covered enough in landscape architecture programs? Do you think it should be emphasized more?
It's not really covered very heavily in landscape architecture programs. It's used extensively in geography and forestry and increasingly in business programs. Sometimes in geology, sometimes in engineering. But very rarely in landscape architecture. Harvard, Wisconsin, Illinois, and a few other programs have it but it's not taught extensively at this point.
So people have to take an elective course on the side to pick up some skills. But if you look to the future of landscape architecture I think it's a very valuable skill. I got into it, really, by chance, and I'd suggest that if you don't like it, don't do it. I'd say don't do anything that you're not attracted to.
So within the profession it's safe to say that you could carve out a niche without GIS skills if you're just not interested in that?
Oh, I think that that's done by thousands of professionals. So it's not an essential skill. If you're doing landscape planning at the landscape level, if you're doing detailed designs, around buildings and campuses it may not be necessary at all.
But do you see it becoming more important in the future as far as for landscape architects?
Here's the situation. Landscape architecture teaches someone the ability to design based on information. That's not a skill that's taught in most professions. It's not taught in forestry . . . Design is not taught outside of the arts in virtually any other field. Landscape designers are certainly not taught the ability to lay out landscape plans or geographic plans – it’s a skill that's unique to landscape architecture. So landscape designers, and I'm talking about plant designers, landscape planners, that skill is very valuable in other fields, beyond the traditional role of landscape architecture.
And getting back to GIS I'm just wondering, do you see that there's elements within the software's capabilities that are relatively unknown amongst landscape architects that could be better utilized?
GIS professionals who are landscape architects usually know most about what the GIS technology does. [But] does the rank and file landscape architect know of the power of GIS? No. I think it's limited to a relatively small percentage, probably just the major part of ten percent.