GIS has been waiting for the cloud. GIS was born to exist in the cloud. It will reach its highest potential there.
The Cloud is fertile ground for a GIS. A GIS is about much more than location alone. Its truest value and highest potential are exposed through the capability to perform spatial analysis, model and simulate. In a sense GIS is a 5-speed F1 racer that has been operating in second gear. The cloud has enormous potential to change that, shifting spatial gears and accelerating the wider use of GIS functionality. Buckle up – the ride is about to begin.
The cloud offers potential in many ways. While sheer computation capability is the most obvious advantage, it also neatly compliments spatial software as a service and server approaches for interacting with spatial data. In also enables many users to interact together, exchanging and collaborating with data pertaining to multiple disciplines. This means a process orientation surrounding dynamic applications can be realised.
Accessibility and collaboration
Many spatial information related tasks pertaining to emergency, military, enviromental, transport, marine, business and other applications inherently involve numerous sources of information and people. They are integrated in nature and dependent upon unique domain knowledge and expertise, even consumer applications. The cloud can be seen as an approach that stimulates collaboration and data sharing through interfaces that are accessible anywhere and by many. Data created locally is distributed and published into the cloud, making it available for use. Most GIS projects benefit greatly when project participants can see, use and create results, placing them back into a common pool.
Interaction and geoprocessing
The cloud offers huge potential for capitalizing upon the spatial analysis, modeling and simulation functions of a GIS. Suddenly the interaction between data and users is not restricted to iterative exchanges, but more fully embodies dynamic exchanges. At the process level this will have major impacts. For example, consider the case of a hydrology application where several people with application programming interfaces (API), and from different domains of expertise, can all be interacting with a process that is connected to their knowledge. Within a cloud, not only do there exchanges interact, but reactions and new data may be created triggering new impacts within nearby processes both directly and indirectly. The cloud computes too.
The size of many GIS data sets and their geoprocessing needs simply cannot be accomplished easily on the desktop. Higher resolution remotely sensed images are about to be available and at much higher frequency due to improved re-visitation times. These images, when coupled to automated image analysis tools existing within a cloud framework or connected to it, will mean higher through-put rates for processing of large quantities of imagery. Lidar and digital terrain models are other examples of data that are not easily processed on the desktop. Highly detailed biological data sets are similarly large. We do not need to have the data in our personal computer to process it.
Users can be distributed anywhere around the world and connecting to projects and tasks of a spatial nature, working on independent variables and problems within larger processes. And, they will also have the opportunity to share APIs and learn from each other as they optimize their routines and work. Interoperability will drive and support this. Everyone does not have to use the same format, instead, interfaces which can dynamically process spatial data and information flows in an interoperable way will be important contributors to making the environment a reality. The ability to interact with the data, through API, is unique to the cloud.
The static to gaming and simulation link
The interactive and dynamic processing of spatial information is important to understand within a cloud environment. It encapsulates higher levels of real data existence and the character of the processes they support. Just as we try to represent reality in map representations, the cloud will enable higher representations of reality where dynamic processes are arising and in 3D space.
Because a higher level of representation can be present in the cloud, using 3D spatial data, it is natural that gaming and simulation are exposed through it. Have you noticed the shift towards photo-realism already? This comes about through the APIs associated with tools originating in the visualisation sector. But I think it happens for another reason.
Clouds are not inherently secure. If they are real clouds, not just utility computing environments, then the API interfaces and data accessibility aspects are integral to their success. They will be for spatial data – the more data revealed, the greater the likelihood the underlying processes will be understood and workable with benefits arising. Clouds like freedom. Sure we can create clouds in our own businesses, but projects like digital cities, military, business and transportation through regions will reach optimised efficiency and effectiveness through openness. This obviously creates questions asrising around data ownership, privacy and other factors. So – how do we get around that? It is like clouds will have in-built angst to be successful.
One way is to shift reality to a game situation. Clouds operating within frameworks of simulation maintain anonymous participation. For example, military applications involving spatial data could be highly valuable if connected to digital city models, real-time transportation networks, scheduling networks and other APIs related to strategic objectives and so on. How do you solve an unknown biological problem in the middle of Europe? You create a game of it, model it and expose it across terrains involving people, infrastructure and knowledge pertaining to it. I can think of lots of applications involving terrorism, disease, energy, pollution, security, business and others that would benefit from cloud/GIS applications.
But to make it seem real means we need people with APIs accessing the processes and entering the cloud to participate in it, live it and play within it. I might add that building information modeling fits in here. Not all cloud projects will be games of course, and there will be a need for people who can develop projects within the cloud, engage people and resources and understand dynamics enough so outcomes can be realised. GIS data will be a fundamental foundation to making this a reality – in military, environment, medical and other applications.
Desktop or cloud?
The desktop GIS is not going away, but much of what we need to be doing with a GIS will be web oriented. Recent cases of Blackberry’s, Google mail and other cloud like services going down has brought some skepticism, especially where needed and emergency services are involved. But we may need to think of clouds as coming in different shapes and styles – nimbus, cirrus, alto-stratus and so on!
For example, there does not have to be one big cloud existing. Several users (2-1000’s) can create a cloud. If you and I set up a GIS server and tell our friends and they tell 5 friends, then we could have quite a few people in our own cloud – albeit a little one. It would move from being a network to a cloud once we give all these people the tools to interact with the servers. GIS began this process a little while ago. If this cloud connects to another one around the world, then it grows. And so and so on. Thus we do not have to think of clouds solely at the Google, Microsoft or Amazon level. Can a big cloud exist but be less useful then another? Yes. Does size matter? Maybe. Knowledge certainly does. Interoperability does. Quality of data would. There will be good clouds, and bad clouds.
What about CAD in the cloud?
I think CAD can experience the same benefits as a GIS within the cloud. Although, I believe it will be particularly useful where the joint CAD/GIS tool sets (and problems) are involved. The relationship of design to space and the processes connecting CAD designs to geographic areas and elements are particularly interesting to consider from a cloud perspective. The area of parametric design is also interesting for design/geospatial applications. What is the relationship of infrastructure to bio-regions, for example. How does building design affect environmental health? What is the relationship of population growth to transportation corridors and buildings in 3D space? How does local energy generation relate to building and sustainable communities? These are some examples of questions involving CAD/GIS combined toolsets.
GIS will become much more exposed, available and useful within the cloud. We will begin to see more of the power of GIS spatial analysis, modeling and simulation within a cloud environment characterized by high quality spatial data, well developed GIS APIs, domain knowledge and governance that supports them. A very high need for cloud – GIS education and computation strategies will be required to enable this work and vision to reach its potential.
The idea of cloud computing is that software can be delivered online as a service, accessing all functionality online, without the need to service or directly control the underlying technology infrastructure. Previously this was known as an application service provider, and there are plenty of successful examples with probably the most prominent mainstream example of Salesforce. Within the geospatial community, there are plenty of good examples of software as a service, with Microsoft’s Virtual Earth and ESRI’s ArcGIS Online Services as good examples.
Some of the rudimentary geospatial tools have been online as a service for quite some time, starting with MapQuest maps and directions. The sophistication of these online offerings has increased over time, and individual organizations have been able to create their own web map services for some time. There remains the distinction that full GIS functionality is a more complex and specialized task with a relatively small audience of specialists with the primary need to author data.
I realize the existence of the debate about “professional” versus “amateur” users, and I just don’t get the irritation. The fact is that very few users of geospatial applications will use expert features for their work. There may be just 5% of the overall user base of a Google Earth or similar application that will need and want extended tool sets for data creation. Making the case for professional versus amateur tools is really just a distinguisher of market size and purpose, instead of a “diss” against those that develop web-based systems and tools.
If the debate is mostly purpose-based, then I think it’s far off base. I suppose another position in the debate is a question along the lines of, “Web-based tools can do all that GIS tools can do, so why would you ever need a desktop tool set again?” I acknowledge that a fair amount of the functions of the professional tools can be satisfied with online tools, but I simply don’t see why you’d want to go online for everything, particularly in the enterprise space. A great deal of control is still needed to accurately conduct spatial analysis or create map products, and I don’t see those tasks going online completely for some time to come.
Control Controls the Debate
Predominantly geospatial capabilities are purchased by organizations, and by companies of such a size and complexity that they feel they must control these systems, particularly when they’re of a critical nature to operations and/or contain proprietary information that must be kept from competitors. I also wonder how long performance, security and reliable accessibility issues will exist for Internet-based applications. I see each three of these issues as considerable barriers for complete cloud adoption, although I’m aware that good progress is being made on the Internet infrastructure front. Power outages do happen, and we’re facing issues of increasing natural events. Simply bringing up the disaster scenario is enough to end the discussion of going solely online in many user circles.
There exists a tool for purpose issue that isn’t easily solved for all delivery mechanisms. Geospatial technologies seem destined for parallel development on desktop, server, web and mobile platforms for some time to come. Geospatial applications will certainly exists on the cloud, and they may become one of the primary markets in terms of the number of users that regularly utilize them. But parallel platforms will exist that are best tuned to different user communities.