All-Source Track and Identity Fuser (ATIF)

Our All-source Track and Identity Fuser (ATIF) automates the process of associating and fusing target data tracks from different sources to understand combat capabilities for enhanced situational awareness.

BAE Systems’ advanced All-Source Track and Identity Fuser (ATIF) technology is a data fusion system designed to detect, process, and report on adversary combat capabilities across surface, ground, air, missile, maritime, and space domains. Initially developed for the Defense Advanced Research Projects Agency (DARPA) to enhance U.S. national security readiness, escalated development and integration of the ATIF software has proceeded under several programs for the U.S. Department of Defense (USDoD), the Department of Homeland Security (DHS), and various intelligence agencies.

How does it work?

ATIF uses artificial intelligence (AI) and autonomy technology to enable tracking and identification of adversaries and their military or intelligence resources by fusing multiple intelligence (Multi-INT) sources, including radar moving target indicators (MTI), image intelligence (IMINT), signals intelligence (SIGINT), video, and other sources of kinematic and classification data. The fuser receives contact or track report data inputs and uses it to produce “all-source tracks” output by associating input tracks with corresponding kinematic, attributed state information, and pedigree. ATIF software has been implemented and deployed in Single Platform, Multi-Platform, or Enterprise estimation configurations. ATIF’s modular design and enabling software suite of capabilities enables rapid adaptation and integration into customer products.

Key benefits of the ATIF technology

From a tracking and data fusion perspective, ATIF has three primary benefits.

First, ATIF improves track continuity. This means that it helps users maintain target track longer by associating contacts and track fragments from multiple sources. This substantially improves decision-makers’ situational estimation by reducing the number of track fragments that an analyst needs to interpret.
Second, ATIF improves kinematic state estimates. For example, MTI-based mover tracks can be provided with target classification information by association to SIGINT or IMINT tracks that are far easier to classify. Estimates of their kinematic state – their measured motion – are also improved by fusion of kinematic information across source and over time. For example, SIGINT tracks can be associated with MTI and IMINT tracks, which significantly improves knowledge of where the SIGINT intercept originated.
Third, ATIF improves target identification. Target classifications improve through the fusion of classification information across intelligence sources and over time. ATIF has a flexible attribution fusion system for using identification features, types, behavior models, and machine learning to create fused identification.

Evolving development for an expanded range of uses

Originally, BAE Systems applied our ATIF technology exclusively to “track-to-track” fusion problems in the ground tracking domain. Today, however, ATIF’s data fusion capabilities have evolved and are being applied to both report-to-track fusion and track-to-track fusion problems. This evolution has expanded its applicability, which significantly increases the practical efficacy of ATIF in surface, ground, air, missile, and space domains for a broad range of defense, intelligence, security, and other areas of government responsibility.

Discover related products that depend on ATIF technologies

Insight – multi-intelligence sensor fusion
CONSENSUS

Discover our related intelligence and security fusion technologies:

Life at BAE Systems

If you want a challenging career creating next-generation technologies that solve complex problems, the BAE Systems FAST Labs™ team may be for you. We are always interested in intelligent, talented creative thinkers who understand both intelligence technologies and how the DoD uses them. Whether you are an experienced intelligence professional or a college grad with a degree in math, engineering, or artificial neural networks, learn more here.