Walking in the Clouds: Unique GIS course harnesses key IT trend

Wednesday, May 16, 2012 - 2:34pm

Unique GIS course harnesses key trend in information technology

By Angela Rogers 




Frank Hardisty, assistant director of the GeoVISTA Center and faculty member for the online MGIS program received an Amazon Web Services in Education Grant to support his first-of-its-kind cloud GIS teaching and cloud computing research, placing Penn State among the first in the world to offer a course in Cloud and Server GIS. The development team for Geography 897C also includes Geography Instructors Sterling Quinn and Ryan Baxter.


“We have wanted to teach a GIS server class for years. However, we were always stopped by the fact that students cannot have administrative control over Penn State servers— for good reasons,” Hardisty notes.


By using Amazon's cloud computing services, students can set up their own (virtual) servers. 




Hardisty and Quinn announced the cloud GIS course at the 2011 Esri Education User Conference.


 What is cloud computing? According to the National Institute of Standards and Technology (NIST):

Cloud computing is a model for enabling ubiquitous, convenient, on-demand network access to a shared pool of configurable computing resources (e.g., networks, servers, storage, applications, and services) that can be rapidly provisioned and released with minimal management effort or service provider interaction. This cloud model is composed of five essential characteristics, three service models, and four deployment models. 


For more details, download the NIST Special Publication PDF or visit the course: https://www.e-education.psu.edu/cloudGIS/


The graduate-level open course uses cloud computing resources from leading GIS cloud vendors like Amazon, Google, and Esri ArcGIS Server, as well as some up-and-coming outfits like OpenGeo and CartoDB. “The course helps prepare students for a career that is going to increasingly rely on cloud computing, in which they must be able to answer the question: ‘Should we rely on cloud computing for problem X?’ where X is a problem relating to the storage, analysis, and visualization of spatial data,” Hardisty says.


In the course, students with no prior programming or server administration experience were able to develop prototype web mapping and spatial analysis services in the cloud.


“It’s a good laboratory for learning about using and administering a server,” Quinn notes.


“Students in the course use cloud computing services to create new GIS web applications,” Hardisty explains. “In a regular online course, student use their own computers to complete course assignments. In Cloud and Server GIS, students use remote computers to complete assignments. So, for example, one student was occasionally in Brazil during the class, and he would connect to computers in Virginia to get work done.”


“Cloud and Server GIS was an amazing, fun and challenging course.  The professors were very knowledgeable and worked closely with all the students to solve problems or provide suggestions on other avenues to pursue. The course was a great opportunity to learn about cloud technology and be exposed to solutions outside the Esri framework,” notes Leslie Zolman, an MGIS grad from Helena, Montana. “With the advancements in cloud technology all GIS professionals need an understanding of geo cloud technologies.”


The practical application is to make GIS more powerful and easier to use, allowing anyone to rent a supercomputer for an hour to conduct a complex geographic analysis.


Hardisty elaborates: “Cloud computers that are very close in power to the PC under your desk or the laptop you carry around cost a few thousands of dollars a year. What's beautiful is that you don't have to rent them for the year, you can rent them by the hour. So, considering that there are 8,760 hours in a year, those same few thousand dollars could also theoretically create a supercomputer with thousands of nodes for an hour. This could be handy if you were trying to do something big and decomposable and you didn't want to wait a year. Many geographic analysis problems are decomposable, so the fit with geography ought to result in more powerful geographic analysis methods becoming practical.”


“Look at what has happened with mapping: Google Maps and similar services have made maps more widely accessible than ever before,” Hardisty remarks. “I see a similar revolution happening with spatial analysis applications.”


The Amazon cloud now has almost 500,000 individual computers in its various data centers. Anyone can access a set of them, for a fee. 


Hardisty notes that cloud computing is one of the big changes happening now in information technology more broadly, and GIS should benefit from it. “Our students will be well served by familiarity with these technologies and concepts as more and more organizations large and small are using cloud computing to cut costs and provide new services.”  


See: Government cloud computing: The good, the bad and the ugly 

Hardisty is also investigating how Cloud Server GIS can be used for research.