Activity-based intel speeds decisions

Demonstration produces targeting quality accuracy in seconds

Montage of world map with tablet and smartphone

Raytheon Intelligence & Space puts Activity-Based Intelligence to the test in a demonstration for Tactical Intelligence Targeting Access Node, TITAN is a tactical ground station that is a forward-deployed system designed to receive, process and disseminate data.

When military commanders need to take action during a conflict, they need the right battlefield data at the right time to make the right decisions.

A team at Raytheon Intelligence & Space, a Raytheon Technologies business, put Activity-Based Intelligence to the test in a demonstration for Tactical Intelligence Targeting Access Node, or TITAN. As a tactical ground station, which is a forward-deployed system designed to receive, process and disseminate data, TITAN sifts autonomously through massive amounts of sensor data to find and track potential threats rapidly.

“At any point in time, there is a finite number of intelligence or data sources available,” said David Appel, RI&S vice president for C2 Digital Solutions at Space & C2 Systems. “That presents a challenge because there are always competing priorities. Activity-based Intelligence helps commanders decide where and when to direct those resources to ensure end-users are getting the insights they need to accomplish their missions.”

Activity-Based Intelligence, or ABI, is an analysis methodology that enables rapid integration of data from multiple sources to discover relevant patterns, determine and identify change, and characterize those patterns.

ABI is an increasingly important enabler of the U.S. Department of Defense’s vision for Joint Domain Operations. Through predictive algorithms, ABI focuses resources on the areas with the highest probability of providing information that is useful for decision-making or action. For example, if a system “thinks” an object is doing something of interest, based on that, it can task additional sensor collection and processing to investigate further.

“By applying predictive algorithms and artificial intelligence, we’re able to optimize how we make use of the data sources available to us,” Appel said.

TITAN can connect to the future Joint All Domain Command and Control network to plan and execute operations in a synchronized and streamlined manner. The demonstration used data from five different sensor types in a real-time processing chain with machine-learning models to generate quality data output.

“What used to take hours, now takes seconds,” said Ari Dimitriou, RI&S TITAN chief engineer. “Before, it took hours to get targeting quality accuracy for an entire image. Now we’re looking at real-time targeting quality accuracy for every pixel of an image.”

The RI&S team pulled from technology across the business to improve speed and accuracy for TITAN processing, including 3D point cloud technology. A point cloud is a highly accurate 3D representation of the Earth, produced from a collection of 2D images. Examples are seen in mapping software that enables the ability to view and exploit 3D representations of buildings, cities or the battlefield.

“We developed the capability to produce accurate high-resolution 3D information from satellite imagery,” said Wyatt Sharp, an RI&S principal engineering fellow for Space & C2 Systems. “Throughout 3D extraction, rigorous error propagation is performed, which yields predicted accuracy for every 3D location.”

The extraction process matches all point locations – this could be a few hundred or millions – from a stack of 2D images, bringing them into alignment to compute 3D locations for each point. When predicted accuracies meet targeting requirements, the data can be referred to as a “targetable surface.”

The targetable surface is the foundation 3D layer that rapidly transfers targeting accuracy to any imaging sensor data to enable prosecution of mobile, time-sensitive or fleeting targets. The technology enables a comprehensive understanding of what’s in an image, but it requires trust. At the end of the day, adversaries are moving to machine-level interactions so they can move faster.

“For us to move faster, we’re going to have to trust the machine-learning automation,” Appel said. “ABI and 3D point cloud technology helps us get there. We’re moving toward a highly autonomous, closed-loop sense-making and orchestration approach to an intelligence problem.”

Effort sponsored by the U.S. Government under Other Transaction number W15QKN-17-9-5555 between the Consortium Management Group, Inc., and the Government. The U.S. Government is authorized to reproduce and distribute reprints for Governmental purposes notwithstanding any copyright notation thereon.

The views and conclusions contained herein are those of the authors and should not be interpreted as necessarily representing the official policies or endorsements, either expressed or implied, of the U.S. Government.

Published On: 01/18/2021