A [Short] History of GIS and Why It Matters

Reflecting on the early days of GIS is critical to understanding stagnation in the geospatial market. Recognizing the profound impact of early adopters, government, data requirements, and limited computer capabilities is key to understanding today’s technologies and vendors. With that understanding, we can take action to move the industry forward.?

One way to look back on the geospatial industry is to look at the types of data used to automate geospatial information.?

Raster Data

Raster scanning technology revolutionized map digitization in the early 1980s, automating information layers and bringing maps to computers. The academic landscape architecture community played a pivotal role in its early development. Picture designing a vast park using just a light table, drafting tools, paper maps, and handwritten attribute data. The designer would meticulously trace layers of information (soil type, vegetation, drainage, etc.) on transparent vellum, carefully coloring in areas to create choropleth and linear/polygon maps (e.g., roads, rivers, lakes, paths). Each "layer" shared common tic marks for seamless overlay. Placing the maps on a light table, and aligning them with the tic marks, the analyst would visually identify overlapping areas and transfer them to separate vellum sheets. Calculating the percentage of a certain characteristic within an area was a manual task. Scanning each data layer and superimposing them marked the early stages of Geographic Information Systems (GIS).

These early systems were raster-based. Jack Dangermond emerged from this field and utilized PIOS, a development stemming from Harvard research, at Environmental Systems Research Institute (ESRI). Later, ESRI devised ARCINFO and GRID for their environmental impact statement work. Simultaneously, the government (specifically the US Army for base management) developed GRASS, which remains accessible today. In 1987, this video, narrated by William Shatner (Captain Kirk of Star Trek fame), offers a comprehensive overview of the state of GIS technology. A significant portion of ESRI's ARC/INFO platform originated from GRASS, leading to ESRI being contracted to port GRASS to workstations.

Although a significant advancement, this leap forward resulted in the generation of a substantial dataset. Why employ numerous rasters for a vast area containing only one type of data, such as water or soil type? Thus, the necessity to convert areas with a singular value, such as soil type, ownership, or lakes, emerged as the subsequent breakthrough in GIS. More insights on this topic will be provided in the upcoming section.

Satellites launched during the late 1970s and 80s were exclusively intended for government use. However, in 1999, IKONOS made history as the first commercial satellite designed for remote sensing. Remarkably, this milestone occurred just 24 years ago. Initially, funding for satellite imagery primarily came from governments, primarily for purposes like spying and reconnaissance. As the availability of this data increased, its potential for non-military applications became apparent. Notably, GPS was initiated in 1973, providing context for the timeline. While ESRI shifted its focus to vectors, companies like ERDAS (an ESRI partner), Intergraph, and others concentrated on developing technology for raster data, specifically for remote sensing and image analysis. These companies dealt with extensive datasets required to examine surfaces and their changes. Furthermore, they played a critical role in translating satellite "pictures" into actionable intelligence. It's important to note that there was a clear separation between companies responsible for data collection (satellites, planes, etc.) and the software used to manage and utilize that data.???

Vectors

During the 1980s, ESRI embarked on the task of vectorizing and building connectivity of map data including lines and areas through the implementation of arc/node topology. For a more detailed understanding of their approach, refer to a paper authored by one of the early developers of ARC/INFO. A significant challenge encountered during this time was the creation of polygons, specifically the complex issue of "island polygons". As maps became automated, the importance of coordinate systems grew. To overlay two maps, they needed to be in the same units. Null Island anchored a wealth of data, thus early GIS systems required the ability to translate data between different projections and handle survey data seamlessly.?

In 1986, when I joined ESRI, my role involved crafting application prototypes for potential clients, showcasing their capabilities, and conducting detailed benchmarks. At the time, ESRI was a services company transitioning into selling its technology to organizations seeking to undertake their projects directly. The software was primarily hosted on substantial computers such as PRIME, Digital Equipment's VAX, and IBM, all with hefty price tags. Early adopters of GIS technology included government bodies like Defense, Departments of Natural Resources, major cities, and national governments. Our primary competitor back then was Intergraph, who approached GIS through Computer Aided Design (CAD) technology developed for circuitry modeling. This article provides a glimpse into the very competitive and ever-evolving landscape of early GIS. The cost constraints and hardware limitations during this stage of GIS development were of paramount importance. As Relational Database Management Systems (RDBMS) like Oracle, Informix, and Sybase became more accessible, non-spatial data became increasingly available, transforming the GIS landscape. INFO which was the tabular side of ARC/INFO was a standalone relational database built by Henco. I built the data dictionary in INFO for an ESRI Air Force project when I first was hired at ESRI. We, as well as users, used INFO directly for manipulating data, populating tables, and building relationships between datasets.??

Approaching GIS through the CAD lens, which focuses on vectors and lines, enabled vendors to push the boundaries of network modeling. Autodesk made its mapping market by automating map creation and enhancing feature attributes, while Synercom introduced database integration for coordinates. The competition, particularly for government users, was intense. The primary goal was to automate and manage the fundamental geospatial layers, as depicted in the popular layer diagram of that era. Additionally, automating The Map was a key focus. I spent countless hours working on cartographic outputs. Notable projects, like the Hong Kong Land Administration benchmark, involved geocoding, permit management, and reproducing custom map products. Pure map automation initiatives, such as those undertaken by National Geographic, Michelin, and the Greek Navy (aimed to generate bathymetric maps automatically at various scales) drove map-making technology even further. The advent of database-driven cartography greatly influenced the early development of GIS.

New Platforms/New Competition

In the midst of all this, a small team in New York was diligently working on a mapping solution for desktops. The solution gained rapid popularity and was known as MapInfo, a game-changer for the GIS and map-making industry. ESRI, in response, swiftly released PC ARC/INFO and DAK the following year, while simultaneously working on ArcView 1, which was launched in 1991 to directly compete with MapInfo. Finally, in 1997, ESRI unveiled the much-anticipated ArcView GIS 3, based on ArcObjects, which laid the groundwork for ArcGIS, officially introduced in December 1999.

During the 1990s, a significant transformation took place in the utility sector with the introduction of GIS. Before GIS, utilities relied on network databases provided by IBM or utilized hardwired technologies like SCADA. IBM's Geographic Facilities Information Systems (GFIS) was one such utility system that major electric companies used to manage their networks. However, as ESRI and Intergraph began promoting GIS solutions to utilities, the landscape changed. This more flexible and automated approach encompassed all assets, leading to a shift in the trade organization's name from AM/FM (Automated Mapping/Facilities Mapping) to GITA (Geospatial Information Technology Association) - a change I witnessed firsthand as a board member. During this time, I also developed prototypes for upstream/downstream tracing, device switching, and more using the relational model of ARC/INFO. The market became fiercely competitive as technology innovation found its way into this critical industry.

A groundbreaking paradigm shift emerged in the form of object-oriented programming and feature modeling. This disruptive force was soon embraced by Smallworld, a start-up hailing from England, which swiftly gained momentum throughout Europe. Its inaugural release in 1989 captured the attention of utilities as it modeled and managed features at the object level which is critical to asset management. Meanwhile, ESRI's librarian model managed data via tiles and the coverages/files level. Leveraging the success of the desktop product ArcView GIS 3 which introduced the geodatabase and feature-oriented data management, ESRI introduced ArcGIS in late 1999, a widely utilized product that remains influential to this day.

The release of Google Maps in 2005 marked a pivotal moment in the evolution of mapping technology. It revolutionized the way we interact with maps, bringing them to our desktops and transforming them into a versatile platform. While MapQuest had showcased the potential of geospatial data and mapping tools for consumers back in 1996, Google Maps took it to unprecedented heights. By introducing the concept of mashups for maps, Google transformed mapping into a dynamic and engaging experience, allowing users to visualize spatial data effortlessly. Even today, countless GIS users rely on Google Maps for geocoding and directions. Additionally, the launch of Google Earth in 2005 provided an entirely new perspective on our planet, presenting it as a vibrant and interconnected organism. This powerful tool continues to be utilized for visualizing global data in an intuitive and captivating manner.

Back in 2009, the release of QGIS, an open-source geospatial desktop solution, shook the industry. Vendors quickly adapt by devising new licensing models and add-ons to keep up with this revolutionary development.

Fast forward to today, and we find a robust geospatial open-source community that includes government, vendors, and users. The shift towards an open and free environment for GIS has fundamentally transformed the market, signaling a dramatic change in the industry's landscape.

Conclusion

Today’s GIS market bears the imprints of its past. As we look to the future of geospatial technology, it is the details of its history that inform us. Within this landscape, many participants, including consultants, data companies, and government projects, each contribute their unique essence to this narrative.

This post aims to explore some of the factors that have shaped the current state of the geospatial market. While some of these influences have become obsolete, new ones have emerged, especially with the remarkable advancements in supporting technology compared to just a decade ago. Among these factors, competition plays a prominent role in driving innovation. It stimulates the development of new products, platforms, ideas, data, and ways of thinking, ultimately creating a new world for geospatial. In this new world, location, spatial relationships, and geospatial thinking are deeply ingrained in our work.

Join us in the Geospatial Innovations group to learn, share, and enable the next generation of geospatial solutions.