Improved MIDAS C-UAS Successfully Tested


Aurora Flight Sciences, a Boeing Company, recently completed a project to advance the capabilities of its Modular Intercept Drone Avionics Set (MIDAS) counter-unmanned aircraft system (C-UAS). Aurora engineers designed, implemented, and tested improvements to the drone engagement device (DED) and onboard autonomy, as well as to the speed and maneuverability of the vehicle platform. Using similar test parameters to last spring's demonstration for the Joint Counter-sUAS Office (JCO) and the Army Rapid Capabilities and Critical Technologies Office (RCCTO), MIDAS autonomously defeated 83% of small UAS targets. "Since our successful customer demos last year, we've continued to improve hardware and software systems on MIDAS to ensure we are prepared to meet the needs of future counter-sUAS programs," said Jason Grzywna, director of small UAS programs at Aurora. "The vehicle we used for the most recent tests included enhancements in agility and autonomy, which improved target acquisition, and in the bolos fired by the DED, which proved more effective in completely disabling the target."

How Facial Recognition Tech Made Its Way to the Battlefield in Ukraine


When the Russian warship Moskva sank in the Black Sea south of Ukraine, some 500 crew members were reportedly on board. The Russian state held a big ceremony for the surviving sailors and officers who were on the ship. But, considering Russia's history of being not exactly truthful when it comes to events like this, many people wondered whether these were actual sailors from Moskva. Toler is director of research and training for Bellingcat, the group that specializes in open-source and social media investigations. He used facial recognition software to identify the men in the video through images in Russian social media, and found that most of the men were indeed sailors from Sevastopol, the town the ship was operating out of.

Artificial Intelligence and the Future of War


Consider an alternative history for the war in Ukraine. Intrepid Ukrainian Army units mount an effort to pick off Russian supply convoys. But rather than rely on sporadic air cover, the Russian convoys travel under a blanket of cheap drones. The armed drones carry relatively simple artificial intelligence (AI) that can identify human forms and target them with missiles. The tactic claims many innocent civilians, as the drones kill nearly anyone close enough to the convoys to threaten them with anti-tank weapons.

As Russia Plots Its Next Move, an AI Listens to the Chatter


A radio transmission between several Russian soldiers in Ukraine in early March, captured from an unencrypted channel, reveals panicked and confused comrades retreating after coming under artillery fire. "Vostok, I am Sneg 02. On the highway we have to turn left, fuck," one of the soldiers says in Russian using code names meaning "East" and "Snow 02." No need to move further. Later, a third soldier tries to make contact with another codenamed "South 95": "Yug 95, do you have contact with a senior? The third Russian soldier continues, becoming increasingly agitated: "Get on the radio.

AI Algorithms Could Rapidly Deploy to the Battlefield Under New Initiative


The Pentagon's Joint Artificial Intelligence Center recently started building a joint operating system and integration layer that combatant commands and other military components could eventually use to rapidly make and field artificial intelligence algorithms. This work is one key piece of the center's new Artificial Intelligence and Data Accelerator, or AIDA, JAIC Director Lt. Gen. Michael Groen confirmed this week during the NDIA 2022 Expeditionary Warfare Conference. "AIDA brings us, in small teams, out to the combatant commanders--now, for those of you who have been in combatant commands, or you're familiar with that environment--combat commanders have all of the challenges, all the problems and only some capability, right? And so what we're trying to do from an information advantage perspective is bring them the advantages of good data and good artificial intelligence-generating insights," he explained. Launched last year by Deputy Defense Secretary Kathleen Hicks, AIDA marks a broad initiative to boost data-based decision-making across the military's 11 combatant commands.

Army Buys Artificial Intelligence-Infused Folding Quadcopters For Battlefield Use – The Drive


Skydio says the X2D offers up to 35 minutes of flight time, can operate in both day or night, and features artificial intelligence tools that enable …

Army Buys Artificial Intelligence-Infused Folding Quadcopters For Battlefield Use


Those requirements could have played a part in the selection process for the Army's SRR program, as the service has previously had to halt the use of foreign-made drones. In 2017, the U.S. Army banned the use of all drones made by Chinese drone manufacturer DaJiang Innovations, or DJI, which supplies over half of all drones sold in the U.S. The company eventually created a U.S. government model in an attempt to allay these concerns. Still, in January 2021, the White House signed an Executive Order that instructed executive branch departments and agencies to examine all of their current drone technologies for potential threats and banned drones and drone subsystems from adversary countries defined as Iran, North Korea, Russia, and China. The Associated Press later reported in June 2021 that Pentagon had cleared some Chinese-made DJI drones for government use, but that report was quickly deemed inaccurate by the Department of Defense (DOD). "This report was inaccurate and uncoordinated, and its unauthorized release is currently under review by the department," the DOD said in a statement in response to the AP report.

Machine learning fine-tunes graphene synthesis


Rice University chemists are employing machine learning to fine-tune its flash Joule heating process to make graphene. A flash signifies the creation of graphene from waste. Rice University scientists are using machine learning techniques to streamline the process of synthesizing graphene from waste through flash Joule heating. This flash Joule process has expanded beyond making graphene from various carbon sources, to extracting other materials, like metals, from urban waste. The technique is the same for all of the above: blasting a jolt of high energy through the source material to eliminate all but the desired product.

Ethics, Rules of Engagement, and AI: Neural Narrative Mapping Using Large Transformer Language Models Artificial Intelligence

The problem of determining if a military unit has correctly understood an order and is properly executing on it is one that has bedeviled military planners throughout history. The advent of advanced language models such as OpenAI's GPT-series offers new possibilities for addressing this problem. This paper presents a mechanism to harness the narrative output of large language models and produce diagrams or "maps" of the relationships that are latent in the weights of such models as the GPT-3. The resulting "Neural Narrative Maps" (NNMs), are intended to provide insight into the organization of information, opinion, and belief in the model, which in turn provide means to understand intent and response in the context of physical distance. This paper discusses the problem of mapping information spaces in general, and then presents a concrete implementation of this concept in the context of OpenAI's GPT-3 language model for determining if a subordinate is following a commander's intent in a high-risk situation. The subordinate's locations within the NNM allow a novel capability to evaluate the intent of the subordinate with respect to the commander. We show that is is possible not only to determine if they are nearby in narrative space, but also how they are oriented, and what "trajectory" they are on. Our results show that our method is able to produce high-quality maps, and demonstrate new ways of evaluating intent more generally. N the 1979 motion picture Apocalypse Now, Captain Willard (played by Martin Sheen) is sent on a mission to assassinate Colonel Kurtz (played by Marlon Brando), a highly decorated officer who, in the words of the general authorizing the mission, has gone from "one of the most outstanding officers this country has ever produced" to someone "out there operating without any decent restraint, totally beyond the pale of any acceptable human conduct." The movie explores the paradoxes in war, where some illegal acts are embraced by the command structure, some tolerated, and some are to be terminated, "with extreme prejudice." Willard has to navigate these conflicts as he moves towards Kurtz' compound deep in Cambodia. Apocalypse Now provides an example of the difficulty that any intent-aware system must face in a military context [1]. Not only does the system need to determine if an order is being followed, it should also determine if the order itself is valid, so that the warriors implementing the order are not placed in ethical dilemmas. This is the goal that we attempt to address in this paper, with the concept of Neural Narrative Mapping (NNM). By placing narrative elements at coordinates in a virtual space, we can determine sophisticated relationships between concepts that go well beyond textual comparison.

Distributed Learning With Sparsified Gradient Differences Artificial Intelligence

A very large number of communications are typically required to solve distributed learning tasks, and this critically limits scalability and convergence speed in wireless communications applications. In this paper, we devise a Gradient Descent method with Sparsification and Error Correction (GD-SEC) to improve the communications efficiency in a general worker-server architecture. Motivated by a variety of wireless communications learning scenarios, GD-SEC reduces the number of bits per communication from worker to server with no degradation in the order of the convergence rate. This enables larger-scale model learning without sacrificing convergence or accuracy. At each iteration of GD-SEC, instead of directly transmitting the entire gradient vector, each worker computes the difference between its current gradient and a linear combination of its previously transmitted gradients, and then transmits the sparsified gradient difference to the server. A key feature of GD-SEC is that any given component of the gradient difference vector will not be transmitted if its magnitude is not sufficiently large. An error correction technique is used at each worker to compensate for the error resulting from sparsification. We prove that GD-SEC is guaranteed to converge for strongly convex, convex, and nonconvex optimization problems with the same order of convergence rate as GD. Furthermore, if the objective function is strongly convex, GD-SEC has a fast linear convergence rate. Numerical results not only validate the convergence rate of GD-SEC but also explore the communication bit savings it provides. Given a target accuracy, GD-SEC can significantly reduce the communications load compared to the best existing algorithms without slowing down the optimization process.