Back to news

February 18, 2019 | International, C4ISR

DARPA: Teaching AI Systems to Adapt to Dynamic Environments

Current AI systems excel at tasks defined by rigid rules – such as mastering the board games Go and chess with proficiency surpassing world-class human players. However, AI systems aren't very good at adapting to constantly changing conditions commonly faced by troops in the real world – from reacting to an adversary's surprise actions, to fluctuating weather, to operating in unfamiliar terrain. For AI systems to effectively partner with humans across a spectrum of military applications, intelligent machines need to graduate from closed-world problem solving within confined boundaries to open-world challenges characterized by fluid and novel situations.

To attempt this leap, DARPA today announced the Science of Artificial Intelligence and Learning for Open-world Novelty (SAIL-ON) program. SAIL-ON intends to research and develop the underlying scientific principles and general engineering techniques and algorithms needed to create AI systems that act appropriately and effectively in novel situations that occur in open worlds. The program's goals are to develop scientific principles to quantify and characterize novelty in open-world domains, create AI systems that react to novelty in those domains, and to demonstrate and evaluate these systems in a selected DoD domain. A Proposers Day for interested proposers is scheduled for March 5, 2019, in Arlington, Virginia: https://go.usa.gov/xEUWh

“Imagine if the rules for chess were changed mid-game,” said Ted Senator, program manager in DARPA's Defense Sciences Office. “How would an AI system know if the board had become larger, or if the object of the game was no longer to checkmate your opponent's king but to capture all his pawns? Or what if rooks could now move like bishops? Would the AI be able to figure out what had changed and be able to adapt to it?”

Existing AI systems become ineffective and are unable to adapt when something significant and unexpected occurs. Unlike people, who recognize new experiences and adjust their behavior accordingly, machines continue to apply outmoded techniques until they are retrained.

Given enough data, machines are able to do statistical reasoning well, such as classifying images for face-recognition, Senator said. Another example is DARPA's AI push in self-driving cars in the early 2000s, which led to the current revolution in autonomous vehicles. Thanks to massive amounts of data that include rare-event experiences collected from tens of millions of autonomous miles, self-driving technology is coming into its own. But the available data is specific to generally well-defined environments with known rules of the road.

“It wouldn't be practical to try to generate a similar data set of millions of self-driving miles for military ground systems that travel off-road, in hostile environments and constantly face novel conditions with high stakes, let alone for autonomous military systems operating in the air and on sea,” Senator said.

If successful, SAIL-ON would teach an AI system how to learn and react appropriately without needing to be retrained on a large data set. The program seeks to lay the technical foundation that would empower machines, regardless of the domain, to go through the military OODA loop process themselves – observe the situation, orient to what they observe, decide the best course of action, and then act.

“The first thing an AI system has to do is recognize the world has changed. The second thing it needs to do is characterize how the world changed. The third thing it needs to do is adapt its response appropriately,” Senator said. “The fourth thing, once it learns to adapt, is for it to update its model of the world.”

SAIL-ON will require performers and teams to characterize and quantify types and degrees of novelty in open worlds, to construct software that generates novel situations at distinct levels of a novelty hierarchy in selected domains, and to develop algorithms and systems that are capable of identifying and responding to novelty in multiple open-world domains.

SAIL-ON seeks expertise in multiple subfields of AI, including machine learning, plan recognition, knowledge representation, anomaly detection, fault diagnosis and recovery, probabilistic programming, and others. A Broad Agency Announcement (BAA) solicitation is expected to be posted in the near future and will be available on DARPA's FedBizOpps page: http://go.usa.gov/Dom

https://www.darpa.mil/news-events/2019-02-14

On the same subject

  • Turkey’s New Akinci Drone Is Impressive, But It’s No Substitute For Modern Fighter Jets

    August 27, 2020 | International, Aerospace

    Turkey’s New Akinci Drone Is Impressive, But It’s No Substitute For Modern Fighter Jets

    Paul Iddon The Bayraktar Akinci drone is the most sophisticated built by Turkey to date. However, Ankara cannot count on the Akinci to serve as a substitute for its air force either acquiring or developing a fifth-generation fighter jet sometime in the next decade. The second Akinci (Turkish for ‘raider') prototype completed a 62-minute flight test in August, according to its manufacturer. The Akinci has a 65-feet wingspan, and is designed to have an impressive 24 hours of endurance, a range of over 300 miles, and to fly as high as 40,000 feet. Ukrainian-built AI-450 turboprop engines will power the heavy drones. According to the Turkish press, the drone “has two 450-horsepower engines but can be equipped with 750-horsepower engines or locally made 240-horsepower engines.” The Akinci will carry a variety of weaponry, including the same Smart Micro Munitions (MAM-L) its predecessor the Bayraktar TB2 does in addition to general purpose bombs. It can also reportedly fire Turkish-built Bozdoğan (Merlin) and Gökdoğan (Peregrine) within visual range and beyond visual range air-to-air missiles. The Akinci can even launch Turkish-built Roketsan SOM long-range air-launched cruise missiles that can hit targets up to 150 miles away. All of this makes it a very impressive weapons platform. Perhaps even more impressive are the drone's indigenous radars. “Ankara sees the Akinci as its main aerial vehicle for intelligence-surveillance-target acquisition (ISTAR) and command-control-communication (C3) tasks in the next decade,” wrote Turkish military expert Metin Gurcan. “The drone will be equipped with indigenously developed systems, including a multi-role active electronically scanned array radar, a SAR/GMTI radar, a wide-area surveillance system, electronic warfare, an electronic and signal intelligence suite, and beyond-line-of-sight satellite communications systems, the sum of which makes the Akinci the best-ever ISTAR+C3 asset the Turkish military has had,” he added. This combination of advanced radars and the multitude of weapons the Akinci can carry certainly make it a highly formidable piece of military hardware. “This weapons system could be quite effective in detecting and destroying individual land targets, such as enemy howitzers or mortars, or special equipment like electronic warfare (EW) stations,” noted a Jamestown Foundation analysis. Turkey has made impressive progress in drone production in recent years. Turkish drones have also seen combat and proven themselves formidable opponents. Its Bayraktar TB2s and TAI Anka-S drones devastated Syrian regime ground forces during clashes in Idlib province in February-March 2020. Turkish drones also gave decisive air support to Turkey's ally in the Libyan Civil War and successfully guided airstrikes that assassinated senior PKK leaders over the last two years. Given this record, it's not at all surprising that Turkey is investing in bigger and more advanced drones like the Akinci since it will be well suited for the kind of conflicts Turkey is fighting and will most likely continue to fight for the foreseeable future. While certainly impressive, the Akinci is, nevertheless, no substitute for Turkey acquiring a fifth-generation fighter jet. As Gurcan noted, by building the Akinci, Ankara is “hoping to extend its military reach in the region and compensate, even if fractionally, for the loss of the F-35 new-generation jets.” Turkey got suspended by the United States from the F-35 Joint Strike Fighter program in 2019 for purchasing and taking delivery of Russian S-400 air defense systems. Also, Turkey's fifth-generation fighter jet project, the TAI TF-X, is unlikely to become operational until the 2030s for a variety of reasons. On top of these serious shortcomings, Turkey might even find it difficult to buy 4.5 generation aircraft in the meantime to serve as stopgap fighters that can gradually replace its aging fleet of fourth-generation F-16s and even older F-4s until Ankara can finally acquire a fifth-generation jet. So while the Akinci is an undoubtedly impressive achievement it's not the aircraft Turkey needs to substantially modernize its air force for the decades to come. https://www.forbes.com/sites/pauliddon/2020/08/25/turkeys-new-akinci-drone-looks-impressive-but-its-no-substitute-for-modern-fighter-jets/

  • Timeline on returning Ospreys to flight remains murky, Air Force says

    February 14, 2024 | International, Land

    Timeline on returning Ospreys to flight remains murky, Air Force says

    Air Force Special Operations Command wants to fly its Ospreys again, but doesn't have the information it needs to do so.

  • Loi de programmation militaire : bilan et mise en garde de la Cour des comptes

    May 13, 2022 | International, Other Defence

    Loi de programmation militaire : bilan et mise en garde de la Cour des comptes

    Dans un rapport présenté mercredi 11 mai, la Cour des comptes dresse un bilan de la loi de programmation militaire (LPM) adoptée en 2018. Les crédits alloués aux armées ont bien progressé de 35,9 Md€ en 2019 à 40,9 Md€ en 2022, comme prévu. La Cour alerte toutefois sur le «  risque d'éviction sur les investissements programmés par la LPM qui restent à réaliser  », pour les années à venir. « La réalisation de “l'ambition 2030” annoncée par la LPM est confrontée au double défi de la dégradation des finances publiques à l'issue de la crise sanitaire et de l'accélération de la montée des menaces décrite dans l'actualisation stratégique de 2021 », prévient la Cour. « Le ministère des Armées doit mieux identifier et exploiter les marges de manœuvre qui peuvent se présenter à lui, notamment dans le domaine de la coopération européenne et s'agissant de la définition du périmètre des missions des armées ». Ensemble de la presse du 12 mai

All news