DARPA expanding Insight program for real-time analysis of ISR data
- By Defense Systems Staff
- Dec 03, 2012
The Defense Advanced Research Projects Agency (DARPA) is moving forward with its Insight program to develop an analytical framework for global, integrated intelligence, surveillance, and reconnaissance (ISR) by issuing a Broad Agency Announcement for phase two of the program. Proposals for the $80 million Insight Phase 2 program are due January 24, 2013, with an industry day planned for December 18 in Chantilly, VA.
“To enhance analysts’ ability to more effectively and efficiently process information, the Insight program is developing an adaptable, integrated human-machine Exploitation and Resource Management (E&RM) System,” states the BAA for Insight Phase 2, which is managed by DARPA’s Information Innovation Office. “Phase 1 of the program created the baseline system with an initial focus on counterinsurgency operations. Phase 2 will mature the Phase 1 capabilities and add additional ones to broaden the system’s applicability to expanded mission spaces.”
Through the development of semi- and fully-automated technologies, the E&RM System will provide the following real-time capabilities in direct support of deployed forces: (1) combination, analysis, and exploitation of information across multiple sources, including imaging sensors and non-imaging sensors; (2) efficient management of sensor tasking; and (3) detection and identification of threats through the use of behavioral discovery and prediction algorithms.
Phase 1 of the program focused on using the E&RM System in scenarios involving irregular warfare, with emphasis on counterinsurgency operations (e.g., identify and defeat insurgent networks). It concentrated on prioritizing sensors, computing power and analytical capability in direct support of tactical brigades and battalions within a wide- area security mission framework.
Phase 1 of the program demonstrated: real-time end-to-end source-to-analyst exercise of the entire E&RM System; the ability to ingest information from ground battle command systems; the ability to ingest information from and dynamically interact with space, air and ground-based sources; sequence-neutral (out-of-order) ingestion and processing of information; the ability to persistently store, index, search and retrieve ingested multi-source information, and present it to an analyst; the ability to detect enemy networks; the ability to track multiple vehicles through suspected drop-off and pick-up vignettes using real-time and forensic rewind of high-value target tracks; the ability to provide timely and relevant information to the tactical edge; and virtual replay and sensor simulation capabilities as tools for system integration and test.
Phase 2 objectives include: the maturing of functions and capabilities developed during Phase 1, while broadening the E&RM system’s applicability to expanded mission spaces (new problem sets spanning operational environments and echelons of command); and the enabling, detection and identification of enemy networks by integrating information from all possible sources, including military intelligence repositories, human reporting, and space, air, sea and ground-based sensors.
Specifically, those objectives include:
• Correlation of multi-source data and information across time and space to identify threat network relationships (consisting of entities, locations, devices, vehicles, events, activities, and other objects, structures, and attributes);
• Correlation of multi-source information from national, theater and tactical systems;
• Automation of manual tasks to free analysts to develop bodies of evidence to confirm or deny hypotheses;
• Georegistration of sensor inputs and intelligence products;
• Enhancements over existing real-time forensic capabilities;
• Tools to monitor more named areas of interest and high-value targets;
• Tools to enhance visualization and interaction with the E&RM System;
• Contribution to the intelligence layer of the commander’s common operating picture;
• Capability to publish standards-based products across intelligence networks in support of other organizations;
• Automated search and retrieval capabilities across intelligence networks to discover available information sources for machine and human reasoning;
• Expanded capability to ingest and process inputs from space, air, ground, sea and specialty sensors, including non-traditional ISR sources (e.g., fighter aircraft targeting pods);
• Exploitation of crowd-sourcing, open-source intelligence and intelligence reporting to supported organizations;
• Expanded capability to ingest and process non-sensor inputs, including data, information and products produced by other systems or analysts;
• Ability to receive and disseminate information and products from external systems operating at any classification level;
• Expanded mission space from counterinsurgency to full spectrum operations;
• Applicability and adaptability to varied operational environments, including denied areas, various physical environments (terrain and weather), and human terrain;
• Applicability to networks beyond insurgent networks, (e.g., command and control, mobile missile, fires, air defense);
• Capabilities that scale;
• Increased scale of time, space and network complexity that can be represented and reasoned over;
• Integration of new inputs and their attributes that must be reasoned over;
• High-fidelity results when incorporating very time-latent information (e.g., days, not minutes);
• Dynamic sensor tasking, cross-cueing and handoff capabilities in the face of an ISR aware and responsive adversary;
• Increased efficiency of sensor resource usage;
• Opportunistic collection from sensors that are not directly controlled;
• Ability to communicate information in support of time-sensitive operations; and
• Product dissemination to disadvantaged users at the tactical edge.