Google Earth Engine: Organizing Geospatial Data to Save the Planet

Google Earth Engine uses open access platform to develop content and applications to make their platform the dominant form in geospatial analysis

 

Until Google launched its Google Earth Engine (GEE) platform in 2010, the geospatial analysis industry was dominated by software produced by ESRI. ArcGIS offered a static software interface that required users to load satellite imagery and data as base layers for their analysis. On the other hand, in GEE, Google had an opportunity to lure users to their own platform by leveraging their significant cloud computing capability to instantaneously upload and update these “base layers,” as well as to analyze data within these base layers much faster than could be done in ESRI’s software package. On launching GEE, Google’s challenge was to prove the value of this nascent GEE platform through creating relevant content and applications to convert historical ESRI users, and to further develop software capabilities that would enhance the usability of GEE relative to the sophisticated analysis packages available through ESRI’s software.

Rather than hire teams of engineers to develop these capabilities in-house, Google granted scientists free access to GEE to serve as beta testers. Scientists were encouraged to program new analysis tools that could be used by others, to provide direct feedback to Google on improving the platform, and to come up with and implement different innovative applications for the powerful combination of computing power and geospatial analysis tools[1]. The basic premise was that the scientists “do the core work of building models and interpreting the data, while Google Earth Engine provides computing heft[2].” That computing heft was significant and a draw to users in an of itself; for instance, GEE could process the data of 700,000 satellite images on 10,000 computers in parallel—a task that would have taken 15 years to complete on a single computer[3]. On the content generation side, one of GEE’s first projects was a partnership with the Surui tribe of Brazil to track and keep records of illegal logging on their land[4]. By outsourcing both content generation and the development of the platform, Google was able to keep their Google Earth Engine in-house team relatively small while proving its ability to compete with (or, arguably, to outcompete) ESRI[5]. Thus, in using an open source approach, albeit one to which Google regulated access, the company was rapidly expanded its platform’s applicability, reach, and capabilities, at minimal overhead cost to the company.

Having increased GEE’s legitimacy, content, and software capabilities by recruiting a broad base of scientists as beta testers, GEE’s next medium-tern challenge is to demonstrate the full scope of their platform at a global scale. Thus the GEE team has shifted its outreach efforts to focus on key, large-scale environmental partnerships. For instance, in July 2018, Google announced a partnership with the United Nations to enable governments, non-governmental organizations, and citizens to track progress against the UN’s Sustainable Development Goals[6]. In October 2018, GEE announced a partnership with National Geographic to track the conservation of nature globally by creating a “four-dimensional digital representation of the vital signs of Earth’s natural ecosystems”[7]. In the medium term, GEE’s challenge will be to ensure that the software they develop is available broadly in an open-source manner that allows for continuous improvement by users.

The GEE team should also explore ways to categorize and document the broad capabilities that have been built by their beta testers in the short term to enable other users to better utilize the platform capabilities and content that have already been built.

In the medium term, GEE should also prioritize democratizing the platform so that non-scientists can benefit from its information organization. Maps are intuitive representations of the world, and expert knowledge is not required to understand content such as identifying areas that are at greater risk of flooding in hurricanes.

A couple of questions remain:

  1. Beyond the motivation to ensure that GEE is the first platform that scientists use—in other words, beyond concerns about the competition from ESRI—what are the other motivators that GEE should consider?
  2. Additionally, GEE was the product of Google’s nonprofit arm, Google.org. Are there commercial applications that could allow GEE to begin to generate revenue and would this be in conflict with the purpose of organizing the world’s information and maintaining the open source nature of the product?

Word Count: 836

Works Cited

[1]Regalado, A. (2010, December 3). New Google Earth Engine. Retrieved from https://www.sciencemag.org/news/2010/12/new-google-earth-engine

[2]Galbraith, K. (2017, December 21). Mapping the World’s Problems. The New York Times. Retrieved from https://www.nytimes.com/2015/01/09/business/international/mapping-the-worlds-problems.html

[3]Galbraith, K. (2017, December 21). Mapping the World’s Problems. The New York Times. Retrieved from https://www.nytimes.com/2015/01/09/business/international/mapping-the-worlds-problems.html

[4]D’Onfro, J. (n.d.). How A “Small But Mighty” Team Of Googlers Is Using Maps To Save People And The Planet. Retrieved November 12, 2018, from https://www.businessinsider.com/rebecca-moore-google-earth-outreach-and-engine-2014-10

[5]D’Onfro, J. (n.d.). How A “Small But Mighty” Team Of Googlers Is Using Maps To Save People And The Planet. Retrieved November 12, 2018, from https://www.businessinsider.com/rebecca-moore-google-earth-outreach-and-engine-2014-10

[6]UN Environment and Google announce ground-breaking partnership to protect our planet. (n.d.). Retrieved November 13, 2018, from http://www.unenvironment.org/news-and-stories/press-release/un-environment-and-google-announce-ground-breaking-partnership

[7]Society, N. G. (n.d.). Life on Earth: Creating a Planet in Balance. Retrieved November 12, 2018, from https://www.prnewswire.com/news-releases/life-on-earth-creating-a-planet-in-balance-300724025.html

 

Previous:

Machine Learning: Schneider Electric Positions Itself to Lead Digital Transformation of Infrastructure

Next:

Bracing for impact: the additive effect of Invisalign on manufacturing

2 thoughts on “Google Earth Engine: Organizing Geospatial Data to Save the Planet

  1. Very interesting article! My view is that this can become a commercial product. I can particularly see how this can be useful in my home country, Nigeria, to track the Boko Haram, the terrorist group. This group resides in northern Nigeria, which is rural and has a landmass that is covered with vegetation, which an effective hiding place for this group. Imagine if our military has this software, which can be used to track movement of this group. I think this can play a critical role in identifying their hide-outs and levelling the playing field, in terms of knowledge of the terrain.

  2. I believe Google is already in the best position to collect and make sense of insanely large datasets and this example is a proof of how they can do it very efficiently and effectively. On top of their mission of organizing world’s information and making it accessible for everyone, with GEE they are also providing opportunities to analyze data to take actions to protect the environment. Thinking about your question, I do think that there might be commercial applications of GEE that can be in line with Google’s mission. They do not necessarily need to sell this as a product, they can get advertising revenues as “sponsorships” from large companies (e.g. Starbucks on their sustainable sourcing) in order to cover their operational costs. Moreover, they can increase the use cases of the engine to areas beyond forests, such as security (monitoring territories) or other environmental applications (cleaning oceans).

Leave a comment