SIG MMG 338 Program Series

Archive for the ‘AI / ML’ Category

Data-Centric Exercise Showcases Joint Capabilities, Lethality

Saturday, February 11th, 2023

FORT BRAGG, N.C. — In the midst of a modernization effort that is focused on meeting strategic demands, organizations from across the military services collaborated in a joint exercise to improve capabilities to prepare to combat near-peer threats.

The XVIII Airborne Corps is at the leading edge of building the future force of 2030 and beyond through collaboration and innovation to meet these complex global challenges.

America’s Contingency Corps put that on full display during Scarlet Dragon Oasis, an artificial intelligence and data-centric operation that enabled a live-fire target identification and execution exercise from Jan. 23 to Feb. 3, 2023.

This was the sixth iteration of the Scarlet Dragon Oasis, and is a continuation of Project Convergence, in an ongoing series of training operations since 2020 that solely serve to increase the capabilities of warfighting skills across the joint force.

The XVIII Airborne Corps partnered with U.S. Central Command and several commands spanning across all branches of the Department of Defense to enable data-driven predictive models to inform and drive operations. The exercise was held across multiple states, including North Carolina, Georgia, Utah and Florida. They employed multiple platforms across all domains to hit specific targets in a more precise and effective manner.

Scarlet Dragon also enabled service members from across the joint force to improve interoperability by incorporating NATO allies. The exercise focused on sharing data and processing artificial intelligence to increase capacity for each warfighting function.

“We see this as an opportunity to take our joint partners and continue training to solve problems together,” said U.S. Army Col. Joseph O’Callaghan, XVIII Airborne Corps Fire Support Coordinator. “Partners are a crucial part of our warfighting force.”

The exercise showcased each service’s ability to operate in a multi-domain environment aimed to enhance their joint relationship.

From the U.S. Army using Boeing AH-64 Apache attack helicopters in a sprawling gunnery range to the U.S. Air Force providing Boeing B-52 Stratofortress bombers, Scarlet Dragon demonstrated the accuracy and lethality of targeting using sophisticated technology.

“The Army is able to find the targets with the software, and then passes that information on to us through our joint domain architecture,” said U.S. Air Force Tech Sgt. Matthew Ping, a tactical air control party specialist. “We pass that information onto Air Force fixed-wing aircrafts, to strike the target.”

The U.S. Marine Corps conducted a High Mobility Artillery Rocket System Rapid Infiltration, partnering with the 18th Field Artillery Brigade using their HIMARS.

The sea domain also played a vital part of the Scarlet Dragon Oasis as the U.S. Coast Guard and U.S. Navy demonstrated their ability to control and patrol the waters by using the Tomahawk Land Attack Missile displaying their targeting precision.

This exercise was also the first time the U.S. Space Force was incorporated into the exercise, and they provided key elements to the exercise by incorporating the space domain and space range to harness another realm of data centric warfare.

“We are using artificial intelligence to pull all actionable data from a scan, instead of having an extra delay from having a team measure it out,” said U.S. Air Force Maj. Joseph Hamilton, a tactical air control party officer. “We’re significantly cutting the time it takes from detection until the point of the strike.”

The joint force model’s interoperability is crucial to our nation’s warfighting success and the improvement of our technology lies within the people who are the service themselves.

“The ultimate goal is if one person from any service walks out with an idea or concept from here and applies it to their work, and makes it better, then we have won. Plain and simple.” said O’Callaghan.

By SPC Osvaldo Fuentes

Some photos by SGT Erin Conway

MarshallAI Starts Cooperation with the U.S. Army

Friday, November 18th, 2022

For immediate release

The Finnish AI pioneer MarshallAI has signed a cooperation agreement with the U.S. Army. The cooperation aims to further develop the MarshallAI platform to suit the U.S. Army even better. The current platform can, for example, analyze the environment on behalf of a soldier, which releases resources for the execution of other acute tasks.

MarshallAI and the U.S. Army have signed a one-year cooperation agreement. The cooperation is a continuation of the international xTech Global AI Challenge organized by the U.S. Department of Defense, where MarshallAI was declared the winner in the fall of 2021. The competition aimed to find innovations based on artificial intelligence that can improve the performance of the different organizations within the Department of Defense.

The MarshallAI platform can among other things multiply the number of eyes on a battlefield. A warfighter can automatically get information about the environment, including sounds and activity in the radio frequencies, which can free up human resources and enable more informed real-time decisions. Warfighters need to cope with less worry and stress when artificial intelligence analyzes and produces information for them automatically.

MarshallAI is a Finnish pioneer of automatic machine vision. The development work started in 2014, and the company offers a solution to replicate human sensing without any knowledge of software development. The technology of MarshallAI improves the safety of companies, enhances the effectiveness of the security authorities, and enables smart cities and transport.

marshallai.com

U.S. Department of Defense Awards Accrete Multi-Million Dollar AI Production OT Software Licensing Contract

Thursday, November 17th, 2022

Accrete’s dual-use AI solution for open-source threat detection, Argus, will enhance DoD’s intelligence competitiveness and bolster U.S. national security

NEW YORK, Nov. 17, 2022 — Accrete has secured a five-year multi-million dollar Production Operational Technology (OT) software licensing contract from the U.S. Department of Defense (DoD) for Argus™, Accrete’s highly configurable, dual-use open-source threat detection AI software solution, worth tens of millions of dollars. The exact contract value and deal terms cannot be disclosed per government guidelines.

Argus continuously analyzes the open-source web to predict anomalous and nefarious behavior hidden in plain sight. Argus performs work that would otherwise require thousands of intelligence analysts at an estimated cost of hundreds of millions of dollars a year in areas critical to national security such as supply chain risk management, vulnerability research, intellectual property theft, social media intelligence, insider threat detection, and anti-money laundering. Argus reads, understands, and learns from an exploding universe of dynamic, unstructured data, including news, blogs, think tank publications, strategy papers, social media chatter, financial filings, microprocessor manuals, and binaries in multiple languages. The AI software automatically extracts, normalizes, and maps relationships between entities, models influence, and surfaces behavioral anomalies indicative of potentially illicit activity that are too complex for humans to identify.

“We are at the dawn of an entirely new type of conflict, driven by rapid advances in artificial intelligence. Whereas nuclear proliferation threatened entire populations, AI proliferation individualizes warfare by manipulating targeted groups through digital means. Using AI, bad actors can propagate disinformation to exploit ignorance and weaken civil society causing billions of dollars of damage to the U.S. economy,” commented Prashant Bhuyan, Founder, CEO, and Chairman of Accrete, Inc. “As a prime defense contractor licensing configurable dual-use AI software directly to the DoD, Accrete is in a unique position to scale its business with the Defense, Intelligence, and Special Operations communities, as well as other government and enterprise customers, and bolster national security as well as intelligence competitiveness in previously unimaginable ways.”

In 2020, the Defense Innovation Unit (DIU), which former Defense Secretary Ashton Carter established in collaboration with former Google CEO Eric Schmidt to strengthen U.S. national security by accelerating DoD adoption of commercial technology, awarded Accrete with an Other Transaction Authority (OTA) Prototype contract after Accrete outperformed 65 other companies in a competitive review. The DoD accepted the Argus prototype within 15 months, and Accrete worked closely with the DIU and the DoD to transition the OTA Prototype contract into a first-of-its-kind multi-year Production OT AI software licensing deal.

William Wall, who leads Accrete’s federal sales subsidiary, Accrete AI Government, and formerly served as a U.S. Special Forces Lieutenant Colonel and with the Joint Special Operations Command (JSOC), said, “We are extremely excited to announce this award, especially as this is the first production OTA that our DoD customer has ever awarded. There are very few start-up technology companies that successfully navigate the extremely challenging OTA process through prototype acceptance and transition into a multi-year production OT software licensing agreement. Our success is a testament to the efficacy, scalability, and configurability of our AI technology and, more importantly, to our employees’ relentless focus on providing solutions that are tightly aligned with end-user requirements. As we continue to scale, Accrete looks forward to providing its growing customer base with powerful dual-use AI solutions that not only drive efficiency through the automation of analytical work but also produce mission-critical insights beyond human capacity.”

Zachary Smith, Accrete AI Government’s Program Manager and retired special agent who spent most of his 23-year Air Force career focused on countering human, technical, and cyber-based threats, observed, “Based on Argus’ blazing speed and scale, in seconds, analysts will be able to research and visualize complex relationships that would have taken at least a month of steady research and annotation. Within the framework of the DoD’s acquisition risk review process, Argus will enable supply chain risk analysis to be more efficient and effective while delivering enhanced value at a lower cost per insight.”

Accrete’s continuously learning AI technology enables the templatization of analytical workflows. Templates such as Argus for Threat Detection learn from sparse data and are configurable for use cases including supply chain influence, social media intelligence, reverse engineering, and logistics disruption for both government and commercial customers. In addition to helping the DoD predict bad actors that may be intentionally obscuring identities to influence the supply chain, Argus has also been enabling the U.S. Air Force to reverse engineer binaries from microprocessor manuals to detect vulnerabilities in firmware.

Accrete’s dual-use AI solutions have also been proven in industry, including media and entertainment companies automating talent scouting to predict emerging talent from social chatter; insurance brokers automating lead generation to find the shortest path to the hottest lead; private equity and venture firms automating pitch deck review through smart table extraction; and automobile dealerships automating the writing of marketing content.

MWW 22 – Turbine One Frontline Perception System

Thursday, November 10th, 2022

Turbine One’s tagline is “AI/ML for the Comms-Contested Battlefield” and after learning about what they are doing, that sums it up quite nicely. For those of you unfamiliar with the terms, AI/ML means Artificial Intelligence and Machine Learning. Those both sound like ominous futuristic concepts but Turbine One’s Frontline Perception System (FPS) makes it not only real, but user friendly.

Unlike current systems fielded with vendor-locked software which cannot move between sensors or devices, FPS combines open system architecture across any network, even in in comms-contested environments, and sensor data fusion.

FPS can be deployed on entire enterprise, desktops, laptops and End User Devices whether connected to a network or not. In fact, it’s optimized for use at the operator level.

It deploys algorithms at the tactical edge to characterize sensor data and provide relevant information to the user. The user can also interact with the system, introducing new items of interest without having to code using their AutoML.

Here you can see an example of the AI identifying a gun. Just prior to taking this photo the system was programmed using AutoML to characterize the pistol after identifying it in several images. It was just that simple.

Finally, TurbineOne routinely partners with third parties to deliver ML to the frontlines.

www.turbineone.com

ONYX Remotely Actuated Weapon

Monday, October 24th, 2022

One of the most promising pieces of equipment meant I saw at AUSA is the Remotely Actuated Weapon by ONYX.

RAW is a Platform-Agnostic Modular Lower Receiver allowing the user to Bring Your Own Upper Receiver.

Currently at Technology Readiness Level 6 (TRL6) signifying it has a fully functional prototype or representational model. The RAW itself is less than 4 lbs and can be mounted to static fixtures or air, ground, or maritime platforms including robotic systems.

It can also be used in conjunction with the Onyx X360 gimbal, which provides 360-degree azimuth and 40-degree elevation integrated with intelligent slew-to-cue automation.

Additionally, ONYX is working on image characterization in order to alert the user to what the system is looking at.