AI Program ‘Project Maven’ Gets Extra $100 Million in Omnibus

April 3, 2018 Laura Criste

This analysis was first available to Bloomberg Government subscribers.

The Pentagon is doubling down on artificial intelligence.

Funding for Project Maven, an Army program that applies AI and machine-learning capabilities to warfighting, surged to $131 million in the fiscal 2018 omnibus that President Donald Trump signed on March 23 from $31 million requested.

The Army Research Lab hosted an industry day in October for Project Maven, also known as Algorithmic Warfare Cross-Function Team, or AWCFT, to better understand industry AI capabilities. More than 100 companies attended.

The extra $100 million shows the federal government’s increased focus on AI as a strategic advantage to solving future challenges, and may signal agency satisfaction with the results of the industry day. Annual AI contract obligations have quadrupled in the last 10 years, Bloomberg Government data shows. This trend is likely to continue for the Pentagon, which has included AI in its Third Offset Strategy and made algorithmic warfare a top priority.

How Project Maven Dollars Will Be Spent

The project’s first phase includes developing computer vision algorithms to help analyze full motion video data. Phase two will improve upon phase one by broadening the scope and speeding up the process to analyze data more quickly and comprehensively.

The Army hasn’t released a solicitation or provided an estimated release date for a final request for proposals. The next step for industry is the Army’s AI and machine learning technology demonstration event from April 30 to May 10. Businesses with AI and ML expertise are invited to submit a “technology capability package.” The Army is interested in demonstration technologies that focus on:

  • Computer vision models;
  • New data labeling techniques;
  • Interfaces for interacting with algorithmically produced data;
  • Storage and indexes capabilities for local algorithmically produced data; and
  • Language algorithms that can process the verbal and written word.

Responses were due in December. BGOV clients interested in this project can set an alert on the program to be notified when ARL posts a request for proposals, probably in mid-May after the technology demonstration event. To continue tracking opportunities related to the AWCFT or Project Maven, clients can save this search and set an alert.

(Laura Criste is a federal market analyst with Bloomberg Government.)

If you’re not a client, and would like to see BGOV in action, click here to request a demo.

Senate Prepares to Fill Trump Team’s Management Gaps