FY 2022 Army Investment in Artificial Intelligence and Machine Learning

Key Takeaways

• Army’s budget for programs leveraging artificial intelligence/machine learning peaked at $1.1B in FY 2020.
• The lower FY 2022 budget request reflects changes in large program budgets, not the Army’s level of commitment to the technology.
• Many of the Army’s efforts investigate the use of AI/ML in tandem with robotic or other unmanned platforms.

With the artificial intelligence hype cycle now going full throttle, it is worth looking at what the Department of the Army plans to invest in the 2022 fiscal year. An analysis of the Department of Defense’s Research, Development, Test, and Enhancement (RDT&E) and Procurement budget requests for FY 2022 reveals a total of approximately $969M that the Army plans to spend on programs with an artificial intelligence or machine learning (AI/ML) component.

Army’s FY 2020-2022 AI/ML-related R&D Budget

The data presented here shows that the Army’s planned FY 2022 investment is down from the roughly $1.1B it budgeted in FY 2020. The cause of the drop is not attributable to a decline in the Army’s commitment to leveraging artificial intelligence. Rather, the budget for one program—the Distributed Common Ground System-Army Intelligence—is projected to fall from $205M received in FY 2020 to $93M requested for FY 2022. DCGS-A had such a high budget to begin with that a lower budget request, which often comes with a maturing program, cannot help but affect the perceived size of the market. Smaller declines in other requested program budgets then combined to make up the balance of the FY 2020 to FY 2022 decline.

Army’s Top 5 Largest AI-Related Programs by Budget

As for the other programs, many of these are investigating AI/ML in tandem with robotic or other unmanned platforms. These systems use AI/ML to augment human operators for multiple purposes, including targeting and navigation.

The bar chart below shows the five largest programs with AI/ML-related investments by requested budgets in FY 2022. Summaries of their basic requirements are discussed below the table.

High Performance Computing Modernization Program (HPCMP): The HPCMP addresses the supercomputing requirements of DOD scientists and engineers. In FY 2022, the HPCMP will demonstrate the potential benefits of multiple architectures (scientific, analytics, machine learning, etc.) that incorporate leading-edge processors, accelerators, memory, data I/O (input/output), interconnect, and OS (operating system) capabilities.

DCGS-A INTEL: FY 2022 Base procurement dollars of $58.5M will be used to continue the enhancement of Army’s intelligence, surveillance, and reconnaissance processing, exploitation, and dissemination capabilities to meet new threats and emerging capability needs. Funding supports equipping and training the force, as well as deploying units with the latest DCGS-A software release. FY 2022 Base procurement dollars totaling $29.1M will be used to update and procure components for DCGS-A Fixed Sites and Army, Reserve, and National Guard units, providing fixed and portable system configurations IAW the Army's Equipping Strategy. FY 2022 funding adds advanced analytics and AI/ML capabilities.

Future Unmanned Aircraft System (FUAS)/Future Tactical Unmanned Aircraft System (FTUAS): The FTUAS will be a runway independent group of three unmanned aircraft that provides Brigade Combat Teams with expeditionary, intelligence, surveillance, and reconnaissance capabilities along with improved targeting. FY 2022 plans include awarding contracts for competitive prototypes that continue the development and integration of required components, such as artificial intelligence.

Information and Networking: This project supports basic research to enable intelligent and survivable command, control, communication, computing, and intelligence systems for the future force. In FY 2022, the Army will investigate models and approaches to enable autonomous systems, explore human-in-the-loop and human-on-the-loop machine learning strategies for interoperability and develop concept recognition, explanation, and inferences for downstream analytics to support collaborative decision making.

Sensors for Autonomous Operations and Survivability Technology: This project designs and develops modular and adaptive sensor components, novel embedded processing approaches, innovative threat cueing solutions and novel multi-function sensor payloads integrated with algorithms and machine learning/artificial intelligence tools to provide improved manned and unmanned ground vehicle situational understanding.

For more information see Federal Market Analysis’ new Federal Artificial Intelligence Landscape, 2022 report due out at the end of August.