IARPA wants intel tools that can predict the future
- By Mark Pomerleau
- Jul 06, 2016
With the flood of information pouring in from sensors and other sources, the Intelligence Community is looking for algorithms that not only can sift through all that data but use it to think ahead.
The Intelligence Advanced Research Projects Activity is seeking proposals in a four technical categories the agency is not currently addressing: anticipatory intelligence, analysis, operations and collection.
As computer algorithms and more agile information platforms that utilize cloud computing have emerged, the idea of anticipatory intelligence to help predict and better connect intelligence dots has become a fabric of the Intelligence Community. IARPA is interested in predictive technologies in detection and forecasting of emergent phenomena, automated generation and maintenance of taxonomies analysis and forecasting of rare events, quantitative risk assessments of emerging dual-use technologies, causal inference from observational data, methods for assessing capability and intent to develop weapons of mass destruction and methods for assessing capability and intent to leverage cyber capabilities against U.S. critical infrastructure.
IARPA, in a recent Broad Agency Announcement, states that analysis research, while cutting across diverse technical disciplines have two similarities. They create technologies that can earn the analyst’s trust by reasoning results and they address data uncertainty and provenance explicitly.
In terms of operations, IARPA explained research efforts under this technical area focus on the ability of the IC to operate freely and effectively in hostile and resource-constrained environments. IARPA pointed out that while adversary activity can stymie these operations, challenges in this domain can also stem from technology changes or business practices. Topics of interest under this technical area, among others, include proposals focusing on robotized research methods, new techniques and applications for algebraic statistics and software development for algebraic statistics with an emphasis on user interface, detection, classification, and mitigation of attempts by adversaries to compromise safety and security. This includes, but is not limited to, penetration and manipulation of electronic infrastructure and assurance techniques that take advantage of emergent enterprise architectural constructs, such as software-defined networking, multi-tenancy and virtual hosting.
IARPA aims to improve on intelligence data by developing new sensor and transmission technologies and techniques that can more precisely target desired information. Combining information from multiple sources as a means of improving the quality, reliability and utility of collected information is another of the agency’s aims. This will become increasingly important as image and signals intelligence collected is poised to significantly increase in the future.
IARPA plans to issue multiple awards under the BAA and awards will generally be for a period of performance lasting 12 months or less. Responses to the BAA are due by May 2, 2017.
Mark Pomerleau is a former editorial fellow with GCN and Defense Systems.