Algorithmic Warfare Cross Functional Team (AWCFT)
Abstract
Project Maven is a rapid fielding AI program that augments and automates PED for FMV of TUAVs, Medium Altitude, High Altitude, WAMI ISR, and other platforms in support of the NDS peer/near peer competitor strategy. Aside from imagery, Maven also uses AI to exploit CEM, Maritime, OPIR and PAI. Maven’s AI, deep learning, and computer vision algorithms and insights are used in theater to detect, classify, and track objects within FMV images (e.g., person, vehicle, and weapon) as well as other projects, such as CEM and text-based projects. Maven algorithms increase the intelligence value of ISR and reduce the human burden of screening so analysts can multi-task, so that analysts and operations personnel can more effectively and rapidly react or more efficiently plan, and so that operations personnel can maintain increased situational awareness. Project Maven, as the pathfinder AI initiative for the DoD, continues to invest in critical architecture to support the rapid expansion of AI to other mission areas besides GEOINT. While most military intelligence exploitation systems were designed pre-AI and require specialized integration to enable the insertion of algorithms into the software baseline, Maven has funded multiple means of bridging these gaps. Project Maven’s commercial technology vendors continue to compete for the opportunity to be inserted into existing programs of record. As Maven algorithms increase in capability, the algorithms have moved to the edge (on sensor platforms). Additional details on accomplishments and capabilities are available at a higher classification level.
Document Details
- Document Type
- Accomplishment
- Publication Date
- Oct 01, 2022
- Source ID
- 6d36facb4da01b23d218d2c57dd4a4bf