SAM #DARPA-PS-26-27

DARPA Solicitation for AI-Driven Tactical Decision Support (DISCORD Program)

DARPA is seeking innovative solutions for the DISCORD program, focused on developing AI-native tactical decision support for military commanders. - Government Buyer: - Defense Advanced Research Projects Agency (DARPA), under the Office of the Under Secretary of War for Research and Engineering (OUSW(R&E)), with support from the Chief Digital and Artificial Intelligence Office (CDAO) - Program managed by Eric Davis, DARPA STO, DISCORD Program Office - Products/Services Requested: - Development and integration of AI-native capabilities for rapid tactical option generation - Physics-based reduced order modeling - AI-driven simulation with edge computing - Online model-predictive control and dynamics learning - Recurrence quantification analysis - Integration of live sensor data with game theory - Human-machine teaming and live adaptation experimentation - Asymmetric advantage engine and cognitive AI partners - Integration into live, virtual, and constructive (LVC) proving grounds - Phased evaluations, integration tests, and end-to-end demonstrations - Unique/Notable Requirements: - Solutions must provide diverse, explainable, and adaptable strategies for commanders - Emphasis on real-time, AI-enabled simulation and human-machine collaboration - Controlled Unclassified Information (CUI) environment - Program is part of DARPA's Ender’s Foundry initiative to accelerate AI-enabled simulation for military advantage - Involvement of government teams from IDA and MITRE for evaluation and demonstration - No specific OEMs, vendors, or part numbers are named in the solicitation

Description

The Defense Advanced Research Projects Agency (DARPA) is soliciting innovative proposals that leverage: abstraction of physics based reduced order modeling, AI-driven simulation with edge compute, advances in online model-predictive control and online dynamics learning, recurrence quantification analysis, simulation with live sensor data, and game theory / theory of behavior. The Disruption through Intelligent Strategies, Counter Options, and Resilient Defenses (DISCORD) program aims to develop an Artificial Intelligence (AI)-native approach for rapid generation of tactical options from both live sensor data and high-fidelity simulation. The goal is to provide commanders with a portfolio of diverse strategies that can pivot dynamically in response to evolving conditions. DISCORD is designed to produce explainable outcomes that human operators and commanders can validate, fostering trust in AI-driven decision support. DISCORD is intended to continuously adapt to changes in the operational environment, producing robust strategies that can pivot dynamically in response to evolving conditions. By discovering a wide range of superior and genuinely diverse strategies, rather than just a single optimized solution, the engine will help create true strategic flexibility. Importantly, this flexibility will afford commanders the power and freedom to exercise their own tactical judgment—born of hard-earned experience, training, and operational art—in ways that adversaries cannot poison, stymie, or accurately predict. DISCORD is a 24-month program organized into two 12-month phases. DARPA anticipates selecting multiple performers.The DARPA DISCORD effort implements Ender’s Foundry, one of several warfighting Pace-Setting Projects (PSPs) established by the Department of War. Ender’s Foundry’s intent is to accelerate AI-enabled simulation capabilities to ensure we stay ahead of AI-enabled adversaries.The effort being solicited by this program solicitation (PS) is controlled unclassified information (CUI).

View original listing