Geospatial technology hasn’t been around for very long. This year marks the 40th anniversary for both ESRI and Intergraph, two of the prominent geospatial vendors. In this short time span, many technical hurdles have been addressed and overcome, and GIS continues to expand its adoption in both scientific disciplines as well as mainstream consumer applications. While much of the promise of the technology has been realized, there are still plenty of frontiers yet to be explored.
There is a certain level of maturity for geospatial technology development as many of the frontiers have been scoped and mapped. Ongoing research and development by academic, government and vendor researchers continues at a rather fast pace, but the technical hurdles of these frontiers are considerable. Even though the path to these achievements may still be difficult, it’s important to keep these ultimate goals in sight. Below are short descriptions of several frontiers, with some details on what might become possible if they’re cracked, and rudimentary explanations of the issues that need to be addressed. By no means is this list exhaustive, and comments and suggestions are welcome to round out the list for a more complete road map.
There have been some very intriguing developments in the realm of 3D visualization lately. The mass collections of highly detailed 3D city models, by Microsoft, Google and others, is perhaps the most compelling. Rich data such as these serve to prime the pump for demand for greater 3D capabilities. The increased realism of 3D environments adds considerable awareness and intuitive interaction, adding greater context and improving understanding.
What’s still missing is the seamless movement between visualization of broad geography into realistic detail to include the interiors of buildings. The combination of the modeling outputs of GIS, CAD and BIM is a well-defined industry objective. A broad and international consortium of vendors and users are working together under the umbrella of the Open Geospatial Consortium to address this issue. There are also vendors such as Safe Software that are working to translate and transform this data into a seamless whole.
The interoperability of data formats and models is the first step toward creating seamless models at all scales that also combine the intelligence of both BIM and CAD for the exciting concept of intelligent 3D models.
One of the primary frontiers is the addition of a time element to geospatial data. Temporal GIS incorporates the X, Y and Z dimensions as well as time for a 4D representation that more closely resembles reality. Time incorporates the moving elements of our world to do away with map abstraction and directly model dynamic processes.
Imagine the amount of insight that you’d gain if you could slow down and speed up time to reveal earth processes, and more directly model the movement of Earth’s inhabitants. We already have video games that can mimic and model the physics of our planet. Wouldn’t it be great if we were capturing our world in rich realism, and combining it with other observations about how things move and interact, including weather?
Inroads have been made to display time series animation of spatial data in small geographies, but the move toward a broad scale temporal GIS is a daunting task. The amount of data needed to bring about the vision is massive. At present, one of the world’s largest supercomputers (the Earth Simulator in Japan) is tasked with crunching temporal data sets in order to model environmental change. The idea of a large-scale interactive model with realistic levels of detail, as well as earth system information, would require a machine of similar capacity for even the smallest city model at such high detail.
Geospatial technology lends a great deal of clarity to situations as they unfold. This decade has seen the rise of prominence of geospatial tools for emergency response and military geospatial intelligence because the tools offer ‘situational awareness’ that forms the basis for multi-agency action to mitigate the detrimental effects of an unfolding situation. The ultimate vision for real-time GIS is to incorporate the viewpoints and data of all actors into a common interface in order to have the fastest and most effective response.
The sensor web plays a large role in realizing the real-time GIS vision. Inroads are being made to incorporate live video and environmental sensors within a GIS framework, and to connect these ground inputs with space-based instruments for a deeper and broader understanding of events as they unfold. There are many elements of real-time GIS in place, including tracking capabilities, but the frontier lies in the combination and condensing of multiple feeds as an incident unfolds.
The Semantic Web
The idea of the semantic web is largely the brainchild of Tim Berners-Lee, who is credited with launching the concept of the World Wide Web. The idea is that the Web will evolve as a medium for knowledge exchange to the point where it will understand the requests of both people and machines so that our computers can do more of the tedious work of finding, sharing and combining information. Geospatial data is a key component of this vision, because an understanding of place is a key differentiator for the relevance of data and for enabling machines to understand commerce.
The idea of intelligent agents has been around for some time, harnessing the power of computers to find the most relevant information for us. The primary driver of the semantic web is to streamline and speed our interactions with less time and thought needed. A key component of speeding interactions deals with proximity and the discovery of the most relevant services that are within the closest distance of our curent location. The location of information also plays into our level of trust of that information, which is a critical issue being addressed in the organization of the semantic web.
A great deal of work is underway to add the ‘where’ component to Web-based information. The creation of the semantic web largely rests in the hands of the search engine vendors who will work to connect the catalog of Web information with its location. While physical points of interest have been cataloged effectively, there need to be inroads into metadata tagging and more automated discovery of the location of information.
Imagery Change Detection
With increasing airborne and space-based imaging sensors, the amount of available imagery is expanding exponentially. Making sense of all of this available information is a daunting challenge. While inroads have been made to catalog, store and retrieve vast amounts of imagery, there are still a great many frontiers in the extraction of information from imagery.
Detecting change in multiple images of the same scene taken at different times is a goal that is being pursued by many researchers. Imagery analysis is a discipline that has been pursued by military practitioners long before the idea of GIS, and a great deal of time and resources are being spent to better harness machines for imagery analysis. While there are a great many algorithms and methods for comparing imagery to reveal cahnge, the geospatial toolset needs to evolve to make this process easier and more intuitive.
Spatial reasoning and analysis are an important aspect of geospatial technology, yet most practitioners just scratch the surface of spatial statistics and the map algebra that can reward the analyst with great insight. While the map is being described as the new interface for mobile applications, there’s a big disconnect between maps as interface and the rich understanding that can be had by thinking ‘mapematically.’
There’s a great deal of complexity in analysing spatial information, and the variable nature of location means that a great deal of error can occur if the correct methodology isn’t applied. The rigorous nature of spatial analysis means that it is a complex and time consuming task.
An ongoing issue with geospatial technology as a whole is the lack of adequately trained practitioners to push the limits of what the technology is capable of, and this knowledge gap is the greatest within the complex area of spatial analysis. There’s a need for much more research into the simplification and automation of spatial analysis in order for the geospatial toolset to reveal greater insights.
From time to time, when I get a bit complacent about incremental industry advancement, I have to remind myself of the enormous potential of geospatial technology. Each and every one of the above scientific and technical challenges are achievable within our lifetime. The current state of geospatial practice has come a long way since the tools were conceived, but we’re still not scratching the surface of the amount of insight that can be unleashed with advancements that are yet to come.
Contemporary GIS and Future Directions, Joseph Berry (GeoWorld, Nov. 2006)
The Semantic Web, Tim Berners-Lee (Scientific American, May 2001)
Challenges for GIS in Emergency Preparedness and Response [PDF], ESRI White Paper (May 2000)
What’s the Promise of Intelligent 3D Models?, Spatial Sustain, Oct. 26, 2007
Why is there so little geospatial analysis?, Spatial Sustain, Dec. 7, 2007
Geospatial Analysis – A Comprehensive Guide, de Smith, Goodchild, Longley (2006)
Image Change Detection Algorithms: A Systematic Survey [PDF], Richard J. Radke (2005)