How technology, procurement, and cultural changes are transforming the business model for GEOINT data
By S. Carter Christopher, Ph.D., NGA Foundation GEOINT Group; J. Edward Pickle, Radiant Solutions; and Nathan Frantz, Geospatial Research Laboratory, U.S. Army Corps of Engineers
Creating and maintaining a “foundational map” of the world is no small feat. While the National Geospatial-Intelligence Agency (NGA) has functional management authority to produce Foundation GEOINT for the National System for Geospatial Intelligence (NSG), the creation of Foundation GEOINT is a team sport. That team is expanding to transform Foundation GEOINT and the business processes/models by which it is compiled. New data options such as crowdsourcing, do-it-yourself data collection from drones and other sensors, and machine learning approaches are coming to supplement traditional foundation products with truly on-demand content.
If GEOINT is formally defined as “the exploitation and analysis of imagery and geospatial information to describe, assess, and visually depict physical features and geographically referenced activities on the Earth,” then Foundation GEOINT is the global representation of those physical features in the form of maps, charts, and data. Foundation GEOINT generally describes natural and man-made features of the world as vector data through points, lines, and polygons. Traditionally created through mono- or stereo-imagery extraction, those points, lines, and polygons are then cartographically styled to create products supporting a range of missions, from U.S. Department of Defense (DoD) operational planning, to aeronautical and maritime navigation, to base maps providing geospatial context in intelligence centers around the globe.
While Foundation GEOINT is inclusive of other mapping, charting, and geodesy (MC&G) components such as surveying, elevation modeling, and gravity modeling; Global Navigation Satellite System/GPS monitoring and processing; and satellite and aerial imagery, LiDAR, and radar collection; this article will focus primarily on vector mapping components. Foundation GEOINT is truly the foundation upon which all geospatially referenced intelligence analysis is built: it provides the cartographic backdrop and location referencing that enables imagery, signals, human, and all-source intelligence.
The origins of Foundation GEOINT in the U.S. can be found in the military services. The Army, Navy, and Air Force all performed portions of the nation’s MC&G needs until the Defense Mapping Agency (DMA) was created to consolidate significant portions of this in the 1970s. Responsibilities for defense and intelligence mapping further converged with the creation of the National Imagery and Mapping Agency (NIMA) in 1996 and have remained as such with the creation of NGA in 2003. Foundation GEOINT has evolved during the history of DMA and NGA, shifting from highly disaggregated (functionally, organizationally, and physically) to highly aggregated. Additionally, while technological advances during this period have been revolutionary across all industries, the accuracy and precision advances in remote sensing now make it possible to map the world at sub-meter resolutions from space, with horizontal and vertical error at less than five meters globally. Combined with aerial assets at much higher resolution and accuracy, medium-resolution/accuracy small sats, and low-resolution but spectrally dense government satellites—such as the U.S. Geological Survey’s LandSat 8 and the European Space Agency’s Sentinel-2—it’s clear the U.S. ability to collect and provide Foundation GEOINT is the best it has ever been.
Even with the significant aggregation of responsibilities at NGA, the agency relies on a team of partners to fulfill the Foundation GEOINT mission. U.S. military services, combatant commands, and intelligence agencies provide data for Foundation GEOINT holdings. Additionally, NGA manages international partner agreements for sharing Foundation GEOINT data and products, and none of this would be possible without the talent and technology of commercial partners and service providers. This article discusses the opportunities NGA and its partners are beginning to leverage in technology, business processes, and acquisition to transform Foundation GEOINT production and delivery, with a purposeful shift toward on-demand access to mission-relevant Foundation GEOINT.
- This article is part of USGIF’s 2019 State & Future of GEOINT Report. Download the PDF to view the report in its entirety.
“The future is already here—it’s just not very evenly distributed.” -William Gibson
The emergence of user-generated content in the form of customer-produced and crowdsourced data enhances the geospatial inventory, but simultaneously blurs lines between traditional Foundation GEOINT and non-authoritative data. Imagery and elevation data collected by drones and tactical aerial assets, and crowdsourced data such as OpenStreetMap (OSM), add to the capabilities of users to collect data “on demand” for their areas of interest. This new generation of user-generated content supplies current and locally specific data to the geospatial community.
One such innovation is the Tactical Full Motion Video to 3D (FMV-2-3D) program currently in development at the U.S. Army Engineer Research and Development Center’s Geospatial Research Laboratory. Using structure from motion, researchers created the FMV-2-3D tool to ingest live FMV feeds and produce customer-sourced Foundation GEOINT products within minutes. These point clouds, digital elevation models, and orthomosaics are available in near-real-time for mission planning and execution at the point of need. No longer are remotely placed teams bound by bandwidth to make critical mission decisions with limited or outdated maps. We are experiencing an era in which end users will collect, process, extract, and visualize geospatial products without touching the GEOINT pipeline.
This evolution of GEOINT product creation has the potential to flip dissemination of GEOINT products to a new model by which the ingestion of field-collected data may outpace the foundation products it seeks to enhance. These soldier-deployed GEOINT collection technologies will only intensify growing content delivery challenges. As more content is required to pass to tactical users on stretched connectivity pipelines, soldiers need intelligent support to reduce the load to only what is necessary for successful mission execution. A critical step forward will take a user- and content-centric approach to support disconnected, intermittent, and limited (DIL) environments.
The emergence of crowd-oriented approaches presents more options for production (and consumption) of Foundation GEOINT. Beginning in 2004 with the development of OSM, map-making expanded from a producer-centric paradigm to a user-centric model. The dominion of highly trained geospatial professionals directed by mapping agency priorities became open to users with a range of skillsets and mapping objectives.
Crowdsourced mapping enables near-real-time access to data and cartographic products. Cartographic products take time to update, and traditional production is difficult to track and monitor. Intermediate data and products that may be fit for use are difficult to access until they are made available as final deliverables. Capabilities such as OSM change this paradigm by enabling access to data and maps as content is created, so consumers are able to evaluate, enrich, and use data when they deem fit.
Maturation of Open-Source Alternatives
OSM and similar volunteer mapping efforts would have limited impact on Foundation GEOINT if not for the rise of open-source software. OSM’s global adoption made it a testbed for software, improving data editing and curation. This aspect of OSM accelerated in 2010, when a devastating earthquake in Haiti led to the creation of the Humanitarian OpenStreetMap Team (HOT) and a wave of open-source, collaborative mapping tools. HOT’s mission required real-time engagement among disparate mappers—including the military, NGOs, civil government, and citizens—and spurred creation of real-time web mapping and mobile software tools. Open-source software engineers in the geospatial community substantially improved the ability for OSM projects to distribute map editing tasks across volunteer mappers, while enabling data quality assurance and curation processes.
OSM’s success was not lost on the U.S. government. NGA began a research and development version of OSM’s user-focused mapping, editing, and curation model five years ago, and this has grown to more than 1,000 domestic and international users. The NSG Open Mapping Enclave (NOME) brings the crowd mapping model to the NSG by allowing NGA’s defense, intelligence, and international partners to crowdsource data needs, enabling approved government users to contribute features and datasets in support of campaigns established based on mission requirements. NOME’s promise is to provide users a “workforce” of mapping collaborators who can work on a project remotely to answer common data needs. NOME offers users both individual and collective ownership of data, allowing anyone to lead a campaign and providing new sources of current map data.
Some Foundation GEOINT missions remain difficult to adapt to the crowdsourcing model or to high levels of machine-derived data. Safety of Navigation (SoN) is one such mission. In the U.S., due to statutory restrictions, aeronautical and maritime SoN data at NGA require a tight chain of custody as well as vetting and approval by certified professionals in those fields.
However, new ground is being broken in these domains. NGA is demonstrating how machine learning and automation technologies can help triage and process global notices to mariners. Additionally, OpenSeaMap and open flightmaps platforms use crowdsourcing models to build global, open navigational maps for the maritime and aeronautical communities.
While these platforms present exciting opportunities, there remain significant hurdles to broad adoption of crowdsourced SoN approaches. First, accurately mapping and updating features (especially non-land-based features) that are not clearly visible in satellite imagery is difficult and requires significant amounts of collateral material to locate, identify, and validate features. Second, the crowd for these types of data is significantly smaller than the crowd interested in mapping land-based features. Lastly, SoN is first and foremost about safety. NGA and other national and defense mapping organizations around the world have strict guidelines and timelines for publishing updates to navigation charts.
Bring on the Microservices
The need for crowdsourced mapping services allowing map editing and review across myriad mappers led to the development of software enabling OSM to harness crowds and update map data with unprecedented speed. A new generation of software is making crowdsourced data more accurate, reliable, and fit for Foundation GEOINT: cloud-based software microservices.
The microservice approach to software development improves the architecting of complex software processes. For Foundation GEOINT, a positive characteristic of the microservices approach is that it allows organizations to flexibly evolve software stacks. For example, for crowdsourced data to contribute to Foundation GEOINT, innovations in software microservices enable NOME and OSM to improve data conflation at scale, mapping campaign management and tasking, map editing, map quality assurance, data publication, and more. This is not to negate the legacy strength and staying power of commercial software that enables the majority of foundation mapping today; no doubt traditional commercial applications will remain central to the toolkit of Foundation GEOINT. However, the opportunities presented by open-source software and custom microservices development vastly expand that toolkit and provide workflow opportunities that are unreachable in legacy applications.
New microservices are on the horizon that allow managers to tailor mapping campaign presets and validations. A manager can limit edits mappers might make, allowing only point or certain types of facility collection. These rules can be embedded into map editing software, providing end users a simple, integrated solution. Other microservices are envisioned that allow Web Feature Services (WFS) to stream data such as points of interest from social media and websites, and to copy/paste features and attributes from the WFS to crowdsourced editing environments. By limiting input error, increasing data for enrichment, and conflating many sources automatically, microservices improve fidelity and enrichment of crowdsourced map resources and make them ready for Foundation GEOINT products.
The Machines Are Coming
The ability of cloud-based computer processes to automatically generate high-quality geospatial products on demand and for large areas is among the most transformational aspect of Foundation GEOINT’s future. This is another example of how the future is already here in some places: At the GEOINT 2018 Symposium, NGA Director Robert Cardillo showed how the “Beachfront” shoreline detection process, developed by NGA’s GEOINT Services, can rapidly draw new coastline vectors as new satellite imagery becomes available.
Processes like Beachfront have the potential to take the time and burden of hand digitization off analysts, enabling them to complete geospatial products in a fraction of current times. NGA partners are using cloud technology and distributed computing to automatically extract building footprints, roads, and more. NGA’s ability to procure these rapidly produced, intermediate products quickly and at lower costs will create an “auto-extraction” services market, featuring agile acquisition and shorter production times that can reshape the Foundation GEOINT market.
World of Data
NGA recently underwent a strategic shift from being primarily a data (and product) provider for the defense and intelligence communities to serving as a broker connecting defense and intelligence users with data and products meeting unique mission needs, some of which NGA may not have a hand in creating or quality checking. This is a marked shift, but necessary due to the massive growth in commercial data providers and in geospatial literacy across NGA customers.
What used to be a tightly controlled data environment almost wholly owned and governed by NGA has shifted to one in which coverage, relevancy, and prevalence of commercial data and products has overwhelmed and outstripped NGA’s, or any government’s, mapping ability. Coupled with expectations for data availability, currency, and coverage by a geospatially literate community of customers, it was imperative NGA reassess its value proposition. As a broker, NGA retains its high-accuracy data and product creation mission while embracing the range of commercial, open, and other government sources that meet customer mission needs, then helps customers reach those data.
With innovative contracts along with organizational shifts, NGA is setting the landscape to enable the agency and its customers to get to the right data, product, or service as quickly as possible. NGA is also exploring ways to evolve acquisition approaches to allow for more competition, quicker execution, and smaller taskings. If success from NGA’s GEOINT Pathfinder program is any indication, the agency’s next step may be hyperlocal microtaskings for Foundation GEOINT updates over priority areas.
As a key example, the U.S. Army has examined the ability to merge NSG authoritative datasets with commercial and open-source products to provide on-road route analysis and navigation at the point of need. As the nation’s principle land force, navigation is a key technology enabler that has changed in focus, emphasizing decentralized operations in dense urban environments. These new warfighting paradigms have stressed current Foundation GEOINT products to provide accurate and timely route analysis at a tactical scale.
With unique mission requirements and specific vehicle movement parametrics, a single routing network and topology source has not been a viable production method. Multiple approaches are being examined, focusing on the ability to populate the operational data models and leverage commercial and open datasets simultaneously. While there exists an accurate geospatial layer to build from, detailed route network attribution (e.g., road directionality, turn restrictions) and temporal information (e.g., road closures, traffic) key to accurate navigation are lacking. Using NSG-procured HERE trucking data, NGA’s NOME road topology, and the Army’s existing road networks, a hybrid solution is being assessed to gather a best-of-breed dataset. While some route attributes will be conflated with other operational schemas, the Army is also investing in technology that will seamlessly switch between government, commercial, and open-source routing networks to provide user-based route analysis across the globe.
The Shift to On-Demand Foundation GEOINT
The future is here for customer-driven and crowdsourced contributions to Foundation GEOINT, enhanced by cloud technology and machine learning. Although these approaches are not distributed equally across the enterprise, they are having notable impacts and are proving successful in pockets across NGA and its community of partners. These approaches will continue to scale horizontally and vertically, and they will necessitate (and in many cases may be drivers of) tighter technology integration across the DoD and the Intelligence Community.
As the NSG continues to embrace more collaborative approaches to Foundation GEOINT mapping and leverages more automation, as NGA grows into its broker shoes, and as the ground warfighter (along with sailors and airmen) produces near-real-time Foundation GEOINT products in theater, NGA and the NSG will need to invest in focused shifts to technology architectures. This includes a protected unclassified production, integration, and collaboration environment for Foundation GEOINT that is globally accessible and scalable to support a global user base. Technology changes, however, will only go as far as culture and policy allow. The NSG will need to invest significantly in shaping these three pillars of transformation simultaneously to fully realize the shift to on-demand Foundation GEOINT data and the operational advantage this provides.
Headline Image: Marine Gen. Joseph F. Dunford Jr., chairman of the Joint Chiefs of Staff, speaks to Robert Cardillo, Director of the National Geospatial-Intelligence Agency, Oct. 24, 2016. NGA delivers world-class geospatial intelligence that provides a decisive advantage to policymakers, warfighters, intelligence professionals, and first responders. Photo by Navy Petty Officer 2nd Class Dominique A. Pineiro.