SAIL-ON: studying the "science" of AI that adapts to the unexpected — a federal contract to RTX BBN Technologies (USAspending)
A federal contract under SAIL-ON, a basic-research program on AI systems that can adapt to unexpected, never-trained-for situations ("open-world novelty"). The Department of Defense (DoD) awarded it to RTX BBN Technologies, Inc.
Contract key facts
- RecipientRTX BBN TECHNOLOGIES, INC.
- Contract value$5,202,312 (≈$5.2M)
- Awarding agencyDepartment of Defense
- Awarding sub-agencyDefense Contract Management Agency
- Award typeDEFINITIVE CONTRACT
- Period of performance2019-11-20 〜 2022-06-30
- Contract ID (PIID)HR001120C0022
Contract scope (original)
THE SCIENCE OF ARTIFICIAL INTELLIGENCE AND LEARNING FOR OPEN-WORLD NOVELTY (SAIL-ON) PROGRAM INTENDS TO RESEARCH AND DEVELOP THE UNDERLYING SCIENTIFIC PRINCIPLES, GENERAL ENGINEERING TECHNIQUES, AND ALGORITHMS NEEDED TO CREATE AI SYSTEMS.
Key points
- SAIL-ON researches the principles, methods, and algorithms for AI that adapts to unexpected "novelty"
- "Open-world novelty" = situations or rule changes an AI never saw during training
- A basic-research contract that funds the "science" underpinning AI, not a finished product
- Recipient: RTX BBN Technologies, Inc.; contract type is a definitive contract
- Touches a root issue of AI robustness and reliability — from self-driving to defense
- AI performs well inside its training distribution and becomes abruptly unreliable outside it.
1What the name SAIL-ON unpacks to
SAIL-ON (Science of Artificial Intelligence and Learning for Open-World Novelty) is a program to research and develop the underlying scientific principles, general engineering techniques, and algorithms needed to create AI systems. "Open-world novelty" means situations or rule changes an AI never saw during training.
Many AI systems perform well within the distribution of data they learned from, but become unreliable the moment they step outside it — and confronting that weakness head-on is what SAIL-ON is about.
2When training assumptions break down
3Basic research behind later applications
Viewed more broadly, this is one example of U.S. research investment at the headwaters of advanced AI. Basic research like SAIL-ON tends to become shared groundwork for many later applications — robustness, out-of-distribution detection, continual learning — rather than a single product.
The recipient is a technology firm with a long track record in applied research, and the very pattern of such companies and institutions executing national research programs reflects a defining feature of the U.S. research ecosystem.
4Inside the distribution, and outside it
Most AI performs well within the distribution of its training data. The problem occurs outside it.
AI in operation inevitably meets situations where the premises of training break — a changed environment, unexpected input, new behaviour by an adversary. Whether an AI can detect such novelty itself and adapt while holding performance is a foundational problem for reliability, from autonomous driving to defense systems. The technical content and results of this contract do not appear in the source.
Why it matters
An example of public funding aimed at a root problem in AI — adapting to unseen situations (robustness and generalization). For readers tracking AI robustness, out-of-distribution detection, and continual learning, it is a useful signal of where U.S. cutting-edge AI research investment is headed.
FAQ
What is "open-world novelty"?
What exactly was built under this contract?
Sources (primary)
This article is an independent organization based on the U.S. official spending data below. Verify the exact, latest details with the official source.
- USAspending (award details)
- Contract ID (PIID)：HR001120C0022