Directions Magazine (DM): Has the cloud become commoditized? That is, is it necessary to separate cloud-based geospatial solutions from others (desktop, server, etc.)?
Victoria Kouyoumjian, IT Strategies Architect, Esri (Esri): I think of "commoditization" as "consumerization,” so, has visualized spatial information (or visualized mapping) become a commodity? I believe the answer is yes, given the dozens of "mapping" companies that offer to visualize your spatial data, consumed via a browser, at home or on a mobile device, or in your car, or at a Hertz kiosk. We have arrived there. But qualifying the difference between GIS in the cloud, and visualized spatial data online is an important element. Generating spatial queries and performing analytics, to reveal patterns or identify trends, for example, are not typically equated to basic “dots on map.” Further, the core characteristics of "the cloud" leverage more than simply an online scenario.
For purposes of privacy (a la private cloud) or convenience (personal cloud) or low-barrier to entry (public cloud), it can be necessary to separate cloud-based solutions from your desktop or server or mobile apps. But this again can be due to political requirements (must adhere to …) or cultural mandates (that's not how it's done) or operational limitations (you're off the grid, my friend).
Peter Batty, VP, Ubisense (Ubisense): It is still early days for cloud-based geospatial solutions, so I definitely would not say the cloud is commoditized in this space yet. You can run more or less any traditional GIS application on a machine at Amazon and say (legitimately) that it is “in the cloud.” And you can probably get some benefit in hardware and deployment costs from doing that. And in many cases you can take systems that originally were designed to run in the cloud, and run them in-house. We did this with our Ubisense myWorld product, since among our customer base of large utilities and telecommunications companies, many are not yet comfortable with running applications outside their firewall. So in many contexts it can be meaningless to say that a given system is running in the cloud or not, you could have an archaic traditional GIS designed 10 years ago running on Amazon, and a modern slick Web mapping application running on an in-house server. But there are some more meaningful distinctions, which I’ll discuss in response to the next question.
Phillip O’Doherty, CEO, eSpatial (eSpatial): Whether it’s necessary to differentiate between SaaS, cloud, server or desktop GIS software really depends on your perspective.
If you’re a GIS expert, a scaled-down SaaS-based application can hardly claim to compete on equal terms with a full-function server or desktop GIS solution. But a full-function SaaS-based solution can – and does – compete on equal terms with server- or desktop-based software.
From the perspective of a newcomer to GIS, SaaS-based geospatial solutions are really the only option. The burden of desktop or server software is seen as unmanageable for a useful – but non-core – software application.
If you’re looking at the software and considering budgets, time to implementation, maintenance and upgrades, then full-function SaaS-based GIS software outperforms the desktop- and server-based solutions.
We believe that, far from being the ugly stepsister, SaaS-based GIS software is actually the Cinderella of geospatial solutions: It has the ability to deliver full, expert functionality, and overcome a host of other issues that otherwise provide barriers to adoption.
Brian Wienke, Product Manager, and Sharon Lin, Product Manager for Government Solutions, Accela (Accela): When discussing the cloud, it’s important to remember the three layers to the cloud “pyramid”: Infrastructure as a Service (IaaS), Platform as a Service (PaaS), and Software as a Service (SaaS). Each layer is built atop the other.
The presence of GIS solutions in the cloud is unlikely to make them commodities. A GIS leader that has historically differentiated itself through value-added features and functionality, user interface, customer support, solutions and customizations can continue to do so in the cloud. As long as these differentiators remain in place, its solutions are unlikely to be commoditized whether they reside in the cloud or on-premise. While a user could change cloud IaaS or PaaS providers who offer similar services, it will likely remain easier to do that than to change GIS vendors and data formats, as there still would be training issues, skill sets and product features to consider.
Mladen Stojic, Sr. Vice President, ERDAS (ERDAS): First of all, there are different levels of commoditization. While the cloud name has certainly become commoditized, the actual implementation and take-up has not. When the geospatial industry first gained an interest in cloud computing, software vendors were relatively quick to assess the potential of the cloud and refactor their existing offerings. In terms of the technology adoption lifecycle, software vendors are in the early majority segment right now. On the other hand, the customers are early adopters, and probably a couple of years away from transitioning to early majority. The customers are still resolving the best way to budget for this technology, and they are still facing some concerns about entrusting the cloud with their hardware and data.
There remains a need to separate cloud-based solutions from the more traditional, site-based solutions because some customers will primarily want to use the cloud for storing data, while some will use it for processing. The industry is still trying to figure out the difference between just storing data and actively using the cloud for processing tasks.
DM: How would you differentiate cloud-based solutions in terms of functionality, cost and ease of use?
Esri: I don't typically consider the differentiators in cloud-based solutions as hinged on ease of use, or functionality or cost. I view the differences between traditional computing and cloud computing as based on (at least) four categories:
- Procurement – Traditionally, organizations would buy assets and build the technical architectures needed (CapEx). In the cloud, you're renting a service, which you often pay for through a frictionless and nearly immediate gratification process that can be extremely attractive to organizations where software or hardware procurement can take many months. Of course, this can lead to other organizational challenges with IT departments, as staff circumvent the procurement process to stand-up a pilot project or proof of concept.
- Business Model – Traditionally, organizations would then continue to pay for those fixed assets, as well as overhead, and administration (CapEx). In the cloud, you are renting those assets, ideally paying based only on usage. At first blush, the latter pay-as-you-go seems like the obvious economical choice, but can sometimes be a head-scratcher when performing budgeting and forecasting spending.
- Access – Traditional computing relies on LANs or WANs or client-server architectures. As you know, cloud is accessed through a ubiquitous network that is (hopefully!) always on, always available. Of course, this may not always be the case and is therefore not suitable for everything!
- Technical – Traditional compute models are static (not dynamic, responding to load) and often dedicated, single-tenant applications or systems. Cloud-based solutions are designed to be scalable and elastic, leveraging multi-tenancy to allow cloud adopters to enjoy some level of economies of scale.
Ubisense: In many cases, old or new generation applications can provide the same functionality running in the cloud or in-house. One area where the cloud can provide more differentiation in functionality is in data sharing between many entities. GeoCommons from geoIQ is a cloud-based repository where users can search and access many datasets uploaded by other users. Envista.com provides good capabilities for coordinating road construction between local government, utilities, communications companies, etc. – a cloud-based approach really simplifies the solution to multi-agency problems like this. The cloud is also very powerful when it comes to “elastic” solutions – for example, systems that are designed to share information about power outages or natural disasters, which normally have a very low number of users but which spike dramatically when an event is in progress. This type of scaling can be easily handled using Amazon, but is very hard to handle with most in-house systems.
In general I think there are compelling cost of ownership advantages to cloud-based solutions. In terms of ease of use, many new generation systems have significantly improved ease of use over more complex traditional GIS, and these solutions can often run in the cloud, but there is no intrinsic reason why use is any different in or out of the cloud.
- There doesn’t need to be a fundamental difference in the functionality offered. Each vendor has to decide whether to commit to full-function SaaS-based GIS, or whether they want to offer a scaled-down set of functionality.
- Advantages of SaaS-based GIS over desktop software include the data efficiencies and collaboration options. Being able to upload one dataset and share it across multiple users creates enormous efficiencies – as does the ability to share items like queries and layers.
- By “renting” the software on a subscription basis, you avoid diving into CapEx budgets.
- Eliminating the need to deal with servers or desktop software also brings down staffing costs.
Ease of use:
- Whether an SaaS-based solution is easier to use really depends on the individual software – and the level of expertise of the end-user.
- Because the SaaS model relies on acquiring and servicing many customers, these companies tend to invest in usability. It’s a practical approach: users need to be able to accomplish tasks without extensive training or intervention. We would tend to agree that SaaS-based software, designed as an SaaS offering, will tend to be easier to use than server- or desktop-based software.
- A major advantage of SaaS-based GIS is that all users will have access to the same version, which eliminates usability issues around learning how to use new versions of the software.
Accela: Broadly speaking, GIS features and functionality should be equal in either cloud or on-premise environments. But an SaaS model allows organizations with limited budgets and resources to leverage the technology without investing in servers and other staff or infrastructure. The benefits include anytime access, economies of scale, faster initial deployment of the platform or software, potentially low impact on IT staff, maintenance-free software updates and patches, and immediate access to the newest release.
ERDAS: In terms of functionality, the differentiator we want to leverage for our cloud-based solutions is their ability to fully utilize hardware resources in the cloud to reduce processing time and provide high production throughput. Intergraph | ERDAS is in a unique position to offer that based on the extensive work we’ve done with distributed processing for ERDAS IMAGINE and LPS.
Our cost differentiator for the cloud is the concept of offering cloud-based services as a monthly subscription package that comes complete with everything you need to run the solution on the cloud, including all necessary software licensing, servers, data storage and bandwidth. This offering is implemented via our partner, Skygone Cloud.
A significant ease-of-use differentiator is the streamlining of the data uploading process. Currently, in the geospatial arena, uploading data to the cloud is typically a laborious process that involves physically placing data on hard disks and shipping them back to the cloud provider.
DM: Do cloud-based geospatial solutions lend themselves to better integration with enterprise business intelligence software, so-called GeoBI?
Esri: It seems pre-destined that geospatial and business intelligence would fit hand-in-glove, integrating traditional business data with spatial data to understand relationships, to improve workflows, revenue streams, customer care or response operations. Of course, geo examples include the location of customers, citizens, channel partners, staff, distribution sites, materials, weather and more. As cloud-based access to these geospatial data lowers the barrier to geo-data consumption, organizations can tap into existing databases or systems, integrated BI content or ubiquitous smart devices. With the advent of geospatial cloud offerings and the need for large-scale intelligent analytics (a la "Big Data"), the adoption of GeoBI seems destined to increase significantly, enabled by cloud-based solutions and consumption models, in order for organizations to improve decision making, business competency, and more effective action (not to mention their bottom-line!).
I recently read an article about a burgeoning hiring trend centering around companies that want to mine the seemingly endless bytes of digital data that they collect or have access to consume. The drive to perform low-level analytics, and make sense of the innumerable data points dove-tails nicely in the GeoBI trending topic — and the introduction of The Data Scientist.
Indicative of where we're going?
Perhaps a Geo-data Scientist is next?
Hmm… that sure sounds like GIS.
Ubisense: No. At least in high-end BI applications, often very large amounts of business data are involved, and there are often significant performance challenges in analyzing such large data volumes. If location aspects of the data are stored or analyzed separately in the cloud (assuming the bulk of the BI data are stored in-house), this would be likely to increase the challenges.
eSpatial: SaaS-based geospatial solutions definitely lend themselves to better integration with enterprise business software.
Integration capabilities are key to the successful development of an SaaS product. While it is not part of an initial offering, it is common for SaaS vendors to go a step further and create an API – with the sole purpose of encouraging integration with other vendors’ applications.
As such, integrations with other SaaS-based software are particularly strong – for example, CRM solutions.
Accela: Having geospatial solutions in the cloud is not necessarily better or worse for integration with GeoBI or other kinds of software that may be on-premise in the enterprise or also in the cloud. Certain vendors that provide SaaS-based solutions have made ease of integration a major objective, which can be beneficial. However, Accela has made it a priority to provide seamless integration with third-party applications in both SaaS-hosted and on-premise environments. Accela’s own integration efforts have not been negatively or positively impacted by the cloud.
ERDAS: The real key to better integration is the establishment of bridges between geospatial solutions and enterprise business intelligence software. Once those critical connections are in place and the integration exists, the entire GeoBI solution can be positioned to obtain all the benefits of the cloud. Many CRMs are recognizing the benefits of the geospatial component, making the integration of this technology more of a reality.
Editor's Note: Follow all the news on cloud computing on Twitter @geocloud and at our Cloud Computing Channel on Directions Magazine; subscribe to our weekly Geocloud newsletter or join the Geospatial Cloud Computing Group on LinkedIn.