ENVI Virtual Analytics Symposium: “Geospatial Vision for the Next Decade”

Tammy Owen, Global Director of the Defense and Intelligence business area for L3Harris Geospatial, introduced speakers on the first day of the 6^th annual ENVI Analytics Symposium held August 25-27^th, which for the first time, was held virtually rather than in person in Boulder, Colorado as had been initially planned.

The good news about the virtual event is that they have doubled their attendance this year over last year, and there are individuals from 36 countries participating this year. The theme for the event was “The Geospatial Vision for the Next Decade.”

The merger of Harris Corporation and L3 Technologies took place last summer. According to L3Harris Geospatial general manager, Eric Young, “we’re hitting all financial targets we had established. We obviously didn’t envision Covid-19 so we’ve all had to find new ways to work; with roughly 50,000 employees almost half have been working from home. So far, we haven’t seen any slowdown in our business. We continue solidifying our new identity as a differentiated technology company. We‘re not a platforms company, we don’t build aircraft, ships or ground vehicles, but we build systems that make those platforms better. We’re a technology company and build mission systems, sensors, and technologies that make the platforms more able to accomplish the mission.”

L3Harris has been acting as a prime with regard to small satellites, and last month announced the (disruptive) launch of an end-to-end small sat mission for the U.S. Air Force, a responsive constellation.

“As we create our geospatial vision for the next decade, we’re leveraging over four decades of experience in image science,” said Young. “We have to constantly innovate because our government, business and academic customers demand better insights and decisions derived from geospatial data.”

A lot of barriers to adoption have disappeared as more easy-to-use tools and workflows make advanced analytics available to anyone, as well as machine learning and deep learning advancements.

Keynote speaker, Pete Doucette from USGS Earth Resources and Observation Science (EROS) Center spoke on “Frontiers of Land Chance Science.”

“My goal was to put together a talk that captures my perspectives from having worked across national security and civil mission spaces over a 30 year career in what I call ‘geoinformatics,’” said Doucette. “It captures the breadth of photogrammetry, remote sensing and GIScience. I started by supporting NGA on the federal side and moved to the contracting side. I returned to federal service five years ago, this time on the civil side with USGS, which has provided me a very different perspective on earth observation.”

Doucette believes that land change science has the potential for cross community development on a global scale. Themes of land change include:

• Monitor – detect what, when, where and how much change occurs

Purpose – quantify the time trajectory of change accumulation

• Analyze – understand why change occurred, and its consequences

Purpose: understand causation and impacts of change and its uncertainty

• Predict – what, when, where, how much, and why change may occur

Purpose: Inform decisions to act

“It has been said that data is the new oil.” This phrase describes how industry has shifted from oil barons to data barons. This gives resurgence to data science.

How does this phrase apply to data science?

Data must be refined for data science to harness its value. Data refinement includes:

• Data conditioning
• Calibration harmonization (over space & time)
• Performant data
• Analysis ready data

Once fueled, we can pursue:

• Data driven analytics
• Inference- turning data into knowledge
• Synthesis
• Prediction

When Landsat data became freely available after 2008, downloads of it soared. Business models changed. There was $2.2B in annual global economic benefit by 2011 and it grew to $3.5B in annual global economic benefit by 2017 (contingent valuation method).

Landsat is first and foremost a global survey mission. The Landsat 8 total refresh for the entire planet’s land surface is a 16 day cycle, collecting about 1 TB day using the “goldilocks regime” of moderate resolution (spatial/spectral/temporal). Landsat data is “analysis ready” with consistent atmospheric correction. The goal of Landsat has always been to balance quantity and quality of data collected where you generally have to sacrifice one to increase the other.

“ARD (analysis ready detection) relieves some users from the processing burden,” said Doucette. Another way to make analysis data ready is by chunking it up. The data is provided on a predetermined grid which includes a fixed geometry. ARD also means providing users with metadata and effects of clouds shown so they can focus on where the data is.”

The traditional way to flag change was to compare observations between two dates and compare their reflectance values and set some arbitrary threshold. “If we didn’t have aerial scenes to reference, we are somewhat limited about what we can say of observations of two pixels over time,” noted Doucette. However, if we look at all Landsat observations since 1984, and model the patterns with harmonic regression function to capture seasonal changes then we can determine amplitude changes and if we label and train a machine learning algorithm, we have the basis for characterizing change both seasonally and per class type.

A fundamental philosophy of LCMAP (Land Change Monitoring, Assessment, and Projection) is that in addition to detecting and monitoring land change over time, the drivers of change also need to be interpreted and understood in order to develop robust models that can provide timely projections of land change for decision makers.

“Once we can understand causation by monitoring the past, we are in a better place to make a long term projection of the future.”

Finally, in order to change culture, we must be adaptable, said Doucette. “How we respond to technological change culturally will shape the outcome of our innovations, because there is a big difference between lofty ambitions and actual outcomes.”

Tags: 3D Laser Mapping, air pollution, autonomous vehicles, climate change, cloud, data, ESRI, geospatial, GIS, GNSS, imagery, Infrastructure, intelligence, laser scanner, location, mapping, maps, mobile mapping, NASA, National Geospatial Intelligence Agency, navigation, NOAA, reality modeling, remote sensing, satellite imagery, situational intelligence, small sats, underground mapping, USGS

Categories: 3D Cities, 3D designs, agriculture, Airbus Defense and Space, aircraft tracking, analytics, asset management, banking, Big Data, climate change, cloud, cloud network analytics, Covid-19, crowd source, data, disaster relief, drones, emergency response, field GIS, GEOINT, geomatics, geospatial, geotechnical, GIS, global aircraft surveillance, government, GPS, handhelds, hardware, Hexagon, image-delivery software, integrated GIS solutions, L3Harris, laser radar, LBS, lidar, location based sensor fusion, location based services, location intelligence, mapping, NASA, National Geospatial Intelligence Agency, National Map, NGA, NOAA, OGC, Open Source, OpenGeo, photogrammetry, public safety, remote sensing, resilient cities, satellite based tracking, satellite imagery, sensors, space-based flight tracking, spatial data, telecommunications, transportation, UAS, UAV, UAVs, USGS, utilities, wildfire risk

This entry was posted on Thursday, September 10th, 2020 at 6:09 pm. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.