Notes from the SatSummit 2018

Yating Ru from IFPRI attended the SatSummit 2018, where satellite industry pioneers and international development experts gathered together for conversations on addressing development challenges with the state-of-art remote sensing technology.
The opening plenary of SatSummit 2018

I’m Yating Ru from IFPRI, and I just attended the SatSummit 2018 held from 19-20 September 2018 in Washington, DC – thanks to CGIAR-CSI’s funding support. The summit summons leaders in the satellite industry and experts in international development to gather together and collaborate on addressing critical issues in global development.

With the advent of a big data era, the unprecedently high-resolution data stream for earth observation enables near real-time monitoring of our environment, which lays the foundation of an agile and robust information-based system in policymaking. But there is a premise that the data users understand the data producers and the latter meet the demand of the former. The convention functions as the bridge for an engaging discussion between the two parties. On the first day, current challenges and obstacles lying in between were identified, followed by heated discussions on responding strategies and solutions. On the second day, more hands-on workshops were organized to facilitate the pick-up of advanced technologies in the day-to-day analysis.

Catherin Andrea Alvarez Hernandez from the Francisco José de Caldas District University in Colombia presented how drone-captured remote sensing imagery is used to extract the habitat information using 3D modeling, such as the size of building entrances in Bogota.
Catherin Andrea Alvarez Hernandez from the Francisco José de Caldas District University in Colombia presented how drone-captured remote sensing imagery is used to extract the habitat information using 3D modeling, such as the size of building entrances in Bogota.

The biggest challenge unveiled was the gap between the end product of the satellite industry – that requires highly specialized technical skills – and the needs of the international development experts – who require ready-to-use data for policy analysis. Luckily, many companies and institutions are well aware of the obstacle and have started dealing with it.  For example, the European Space Agency is continuously incubating the Copernicus Open Access Hub for a user-friendly pathway to its Sentinel-1, Sentinel-2, and Sentinel-3 products by providing an attractive visual display and a simple click-and-download choice. During a panel discussion between four major satellite industry leaders –  U.S. Geological Survey, European Space Agency, DigitalGlobe, and Planet – all of them recognized the importance of user end delivery and anticipated the continuous improvement of the user interface. Actually, some third parties have already built intuitive platforms to link the data with the development community like Radiant Earth Foundation Platform. It deployed a most succinct and easiest user interface to conduct basic analysis of data including Landsat, Sentinel, and many high-resolution commercial satellites. I’m very glad to see so many efforts on user end delivery and personally look forward to a powerful platform that incorporates all the data sources and functions used in development research, but easy to use.

A.I. 101 panelists from Development Seed, NVIDIA, DigitalGlobe, and Microsoft

Another topic that keeps coming up is about ground-truthing data, which are the key to bring volumes of data to play their role in the actual research. As we know, satellite data have to be adjusted and validated using ground truthing data before they are applied to automatically infer information for a large area. But collecting ground truthing data is time-consuming and labor-intensive. Many companies have infrastructure and technologies to process the satellite imagery for the derivative product such as super high-resolution land cover and road network, but they lack ground truthing data. On the other hand, the development community has accumulated abundant field data but lacks the storage and computation power to produce the satellite derivative data in dire need. During a session, a program manager on the Bing Maps at Microsoft shouted out “Anyone here in the room has access to African ground truthing data? Please let me know! Let’s talk.” It turned out we CGIAR has rich data gathered by different centers and there may be a potential opportunity for Microsoft and CGIAR-CSI to collaborate. It is foreseeable that more and more partnerships will take shape in sharing and exchanging ground truthing data. And some old data will be brought to life again leveraging on the latest technologies.

Remote sensing in food security presented by Ritvik Sahajpal from University of Maryland

During the hands-on workshop sessions, I had practical training on Google Earth Engine, GBDX, and Radiant Earth Foundation Platform. Taking advantage of cloud storage and cloud computation, these platforms make it possible to complete tasks that were impossible by individual computers. The smooth experience attracts users to shift more local work to the online platforms, which are also in favor of sharing and communicating. We have reasons to believe that as the platforms become more and more user-friendly, the open source geospatial data will become truly open to every person, including those who only have minimal knowledge of geospatial science.

A number of other interesting sessions sprung up as well, such as the application of remote sensing in food security and discussion on policy requirements and privacy protection. As the partnerships between the satellite industry and the development community build a foundation for freely exploring geospatial data in the policy analysis, the growth of thrilling topics can take on an exponential path. Everything is evolving at an accelerated pace. Isn’t it exciting?