Not since the transition from film to digital imagery has the geospatial analysis profession experienced such a rapid transformation
By Damon Brady, Sr. Director, Product Development and Programs, BAE Systems; John Steed, Director, Geospatial Services, Tesla Government; and Anthony Sanchez, Technical Director, Veritone
The exponential growth of sensors, geospatially relevant data, and advanced analytics has created increased opportunities for geospatial analysis and is quickly leveling the asymmetric GEOINT leadership advantages previously owned by few. Not since the transition from film to digital imagery has the geospatial analysis profession experienced such a rapid transformation, and the new “digital universe” of geospatially relevant data continues to expand, incorporating and inspiring new technologies.
The statistics are impressive: According to 2017 market reports from Tauri Group, Earth observation satellite count is growing at a five-year Compound Aggregate Growth Rate of 47 percent and by 2025 we will see more than 750 new “eyes in the sky” providing imagery, videos, and multiple other types of data for analysts to consume. Similarly, the commercial drone population (excluding military and personal drones) is expected to grow at least 50 percent to more than 800,000 in the same timeframe, according to BI Intelligence reports. When combined with the ubiquitous proliferation of Internet of Things sensors, Forbes predicts our digital universe could expand to more than 163 zetabytes by 2025. Already, there is so much information available that significant amounts of data cannot be processed in the traditional processing, exploitation, and dissemination workflows, resulting in potential loss of information dominance and the inability to extract new insights and value.
- This article is part of USGIF’s 2019 State & Future of GEOINT Report. Download the PDF to view the report in its entirety.
Charting a Path to Transformation
Mission leaders are challenging the GEOINT Community to lead the way in this complex new environment. It is a logical ask—after all, one of the best ways to correlate and understand the relationships among different datasets is through images. As a result, GEOINT is evolving beyond providing imagery and geospatial awareness to be the epicenter of data fusion, responsible for the interpretation and visualization of highly disparate, differentiated data streams. While accepting this challenge and driving forward with GEOINT transformation is an exciting prospect, we must also recognize that this magnitude of organizational change is neither easy nor immediate. It is an iterative process that will evolve over time, yielding a result that will likely be substantially different from what was initially envisioned.
Transformation begins at the top, with the effective communication of vision and mission objectives needed to align stakeholders in the pursuit of common goals. These objectives are based on a recognition of our increasingly important role across intelligence and data-driven disciplines. They define an underlying value proposition developed with a realistic understanding of the impact transformational change has on the enterprise’s people, processes, and tools. In order for the community to succeed in our GEOINT transformation, we must employ a holistic, integrated approach to tackle the main issues related to big data, human capital optimization, and the application of advanced technology tools to meet the needs of both the GEOINT and broader Intelligence Communities.
Implementation of an integrated approach is often easier said than done. The reality is there are many challenges and obstacles in the way of achieving transformative success. Tactically, this includes things such as the likely requirement to modify existing workflows and processes in critical areas where operational disruption is not an option, managing issues around operations security and sharing information between organizational silos, and mitigating concerns related to data integrity and provenance. Strategic challenges may be even greater with fundamental business decisions to be made around evolving technology, funding availability, and an ever-changing political landscape.
These challenges are ever-present, and what remains at the end is the need to understand what can be done to make these goals actionable. There are several ways to chart a path, including leveraging business process transformation or organizational change management methods. Given the importance and relevance of technology in the mission, however, it may be most appropriate to approach the challenge from a systems engineering perspective. Application of a systems engineering methodology provides a requirements-driven framework to decompose large systems into their constituent subsystems and components, apply change, and rebuild as an integrated and holistic system. For the diverse and rapidly evolving GEOINT Community, this approach has the benefit of focusing expertise and creating an environment of iterative, actionable, and measurable steps toward achieving transformation goals.
Unlocking Value from Big Data
One of the key areas that can benefit from this approach is big data. The GEOINT discipline is facing immense challenges in keeping up with the amount of data to process, analyze, and disseminate. This is not only because of the sheer quantity of overhead imagery collected, but also a result of the surge in full-motion video, all-source, and unstructured data. The kind of problems the GEOINT Community will need to solve tomorrow are no longer solely based in geospatial context. The addition of open data, surveillance streams, and vast amounts of unstructured text from news sources, social media, and other communications makes new insights available to the GEOINT Community. The challenge is to incorporate this additional information into valuable, relevant, and timely intelligence.
Today, there is simply too much information to process and analyze using yesterday’s systems. As GEOINT has evolved to become a focal point for data fusion, a new demand has emerged for geospatial tools and workflows that creatively solve future unknown problems.
Fortunately, artificial intelligence (AI) and machine learning (ML) can help drive the community
toward achieving these objectives. The rapid rise of AI/ML means that capabilities are becoming commoditized and increasingly prevalent in the commercial space. They are currently providing industries such as advertising, law, medicine, and finance the ability to extract meaning from unstructured sources at scale. Inserting AI/ML technology into the GEOINT Community was inevitable since the data processing requirements call for increased data-driven and automated solutions. Near-term AI/ML solutions are focused on solving problems concerning throughput and productivity. That is, automating repetitive tasks analysts use to search and discover intelligence in real time. This type of application is low-hanging fruit for the technology, but it’s just the tip of the iceberg as the pace of the market moves faster than the pace of innovation.
The approach in the GEOINT Community thus far has been to invest in point solutions that solve narrow use cases as a proof of concept. This is a typical way for government to add most new technology and software. However, the concern is that the cost and capability of AI/ML technology is changing so rapidly that initial investments made become outdated and decrease in value as each day passes. To solve this problem, organizations must think about incorporating AI infrastructures that flexibly scale with the changing technology. A GEOINT future with AI/ML augmented solutions will provide highly adaptable ecosystems of analytic capabilities that are designed and leveraged to cost-effectively solve real-time, mission-critical tasks.
People are Still Essential
Many changes have occurred throughout the GEOINT industry in the last half century, from heads-up digitizing, satellite improvements, and big data management to AI/ML and the movement from using software interfaces to leveraging scripting and coding languages to perform tasks. The pace of change has increased significantly in recent decades, and while a highly debated topic, some researchers predict that computers will achieve basic human cognitive capability within 20 years. This has the potential to automate many of the basic, repetitive tasks currently performed by human analysts.
In the interim, the question remaining is, “How do GEOINT teams keep up with the fast evolution of technology and data while maintaining institutional knowledge and expertise?” It is paramount that the role of seasoned veteran analysts continues to guide our understanding of what is valuable from GEOINT data, but new data and processes must be leveraged to improve the speed of production and to derive new, valuable insights that may not have been previously accessible or conceivable.
There are several areas of potential focus in human capital optimization that will help realize the benefits of GEOINT transformation quickly while maintaining the mission-critical quality and reliability provided by our highly skilled workforce:
- Leverage institutional knowledge: Strong links should be created between teams that configure technology and the analysts who leverage it. Just like a race car is built with the driver in mind, insights from experience can help guide and ensure that technological advancements and adoption evolve in a symbiotic way.
- Train the trainable: Create new methods to incorporate new learning for seasoned analysts. Not all new technologies need to be adopted and used by technophiles. Building capacity for existing experts will add more value to already valuable assets.
- Encourage on-the-job technology research: Providing time for GEOINT analysts to research changes and trends in the industry can lead to a more educated and industrious workforce. The tools of the trade and new options for improving old tasks should be on the minds of workers, and this should be institutionalized from leadership through mid-level management to ensure the GEOINT Community remains competitive.
- Create skills transfer programs: Some new skills (such as tool automation) can be trained to personnel across departments, enabling improved workflow and collaboration throughout different teams and disciplines. Enabling personnel to train skills to others not only reinforces that skill, but also builds cross-team/department cohesiveness and expands capabilities across the organization.
Achieving GEOINT transformation and moving our field forward requires rethinking how workflows have been developed and managed in the past. As leaders communicate strategic vision and mission goals, we can apply proven methodologies to break down traditional information silos and iterate how we apply limited resources toward developing integrated solutions for the future.
A holistic view and willingness to adapt is important to effectively capitalize upon technology advancements that can address today’s big data glut, and advanced technology must be thoughtfully but aggressively applied to achieve better, faster, and more confident operations. Perhaps most importantly, ways in which we acquire, train, and deploy human capital must be re-imagined to focus on speed, assessment, and complex decision-making. The GEOINT transformation will drive new best practices in each of these areas to enable the data fusion needed to maintain our information and intelligence advantage.