2 novembre 2020 | International, Terrestre, C4ISR

How the Army plans to revolutionize tanks with artificial intelligence

Even as the U.S. Army attempts to integrate cutting edge technologies into its operations, many of its platforms remain fundamentally in the 20th century.

Take tanks, for example.

The way tank crews operate their machine has gone essentially unchanged over the last 40 years. At a time when the military is enamored with robotics, artificial intelligence and next generation networks, operating a tank relies entirely on manual inputs from highly trained operators.

“Currently, tank crews use a very manual process to detect, identify and engage targets,” explained Abrams Master Gunner Sgt. 1st Class Dustin Harris. “Tank commanders and gunners are manually slewing, trying to detect targets using their sensors. Once they come across a target they have to manually select the ammunition that they're going to use to service that target, lase the target to get an accurate range to it, and a few other factors.”

The process has to be repeated for each target.

“That can take time,” he added. “Everything is done manually still.”

On the 21st century battlefield, it's an anachronism.

“Army senior leaders recognize that the way the crews in the tank operate is largely analogous to how these things were done 30, 45 years ago,” said Richard Nabors, acting principal deputy for systems and modeling at the DEVCOM C5ISR Center.

“These senior leaders, many of them with extensive technical expertise, recognized that there were opportunities to improve the way that these crews operate,” he added. “So they challenged the Combat Capabilities Development Command, the Armaments Center and the C5ISR Center to look at the problem.”

On Oct. 28, the Army invited reporters to Aberdeen Proving Ground to see their solution: the Advanced Targeting and Lethality Aided System, or ATLAS.

ATLAS uses advanced sensors, machine learning algorithms and a new touchscreen display to automate the process of finding and firing targets, allowing crews to respond to threats faster than ever before.

“The assistance that we're providing to the soldiers will speed up those engagement times [and] allow them to execute multiple targets in the same time that they currently take to execute a single target,” said Dawne Deaver, C5ISR project lead for ATLAS.

At first glance, the ATLAS prototype the Army had set up looked like something out of a Star Wars film, albeit with treads and not easily harpooned legs. The system was installed on a mishmash of systems — a sleek black General Dynamics Griffin I chassis with the Army's Advance Lethality and Accuracy System for Medium Calibur (ALAS-MC) auto-loading 50mm turret stacked on top.

And mounted on top of the turret was a small round Aided Target Recognition (AiTR) sensor — a mid-wave infrared imaging sensor to be more exact. Constantly rotating to scan the battlefield, the sensor almost had a life of its own, not unlike an R2 unit on the back of an X-Wing.

Trailing behind the tank and connected via a series of long black cables was a black M113. For this demonstration, the crew station was located inside the M113, not the tank itself. Cavernous compared to the inside of an Abrams tank, the M113 had three short seats lined up. At the forward-most seat was a touchscreen display and a video game-like controller for operating the tank, while further back computer monitors displayed ATLAS' internal processes.

Of course, ATLAS isn't the tank itself, or even the M113 connected to it. The chassis served as a surrogate for either a future tank, fighting vehicle or even a retrofit of current vehicles, while the turret was an available program being developed by the Armaments Center. The M113 is not really meant to be involved at all, but the Army decided to remotely locate the crew station inside of it for safety concerns during a live fire demonstration expected to take place in the coming weeks. ATLAS, Army officials reminded observers again and again, is agnostic to the chassis or turret it's installed on.

So if ATLAS isn't the tank, what is it?

Roughly speaking, ATLAS is the mounted sensor collecting data, the machine learning algorithm processing that data, and the display/controller that the crew uses to operate the tank.

Here's how it works:

ATLAS starts with the optical sensor mounted on top of the tank. Once activated, the sensor continuously scans the battlefield, feeding that data into a machine learning algorithm that automatically detects threats.

Images of those threats are then sent to a new touchscreen display, the graphical user interface for the tank's intelligent fire control system. The images are lined up vertically on the left side of the screen, with the main part of the display showing what the gun is currently aimed at. Around the edges are a number of different controls for selecting ammunition, fire type, camera settings and more.

By simply touching one of the targets on the left with your finger, the tank automatically swivels its gun, training its sights on the dead center of the selected object. As it does that, the fire control system automatically recommends the appropriate ammo and setting — such as burst or single shot — to respond with, though the user can adjust these as needed.

So with the target in its sights, weapon selected, the operator has a choice: Approve the AI's recommendations and pull the trigger, adjust the settings before responding, or disengage. The entire process from target detection to the pull of the trigger can take just seconds. Once the target is destroyed, the operator can simply touch the screen to select the next target picked up by ATLAS.

In automating what are now manual tasks, the aim of ATLAS is to reduce end-to-end engagement times. Army officials declined to characterize how much faster ATLAS is than a traditional tank crew. However, a demo video shown at Aberdeen Proving Ground claimed ATLAS allows “the operator to engage three targets in the time it now takes to just engage one.”

ATLAS is essentially a marriage between technologies developed by the Army's C5ISR Center and the Armaments Center.

“We are integrating, experimenting and prototyping with technology from C5ISR center — things like advanced EO/IR targeting sensors, aided target algorithms — we're taking those technology products and integrating them with intelligent fire control systems from the Armaments Center to explore efficiencies between those technologies that can basically buy back time for tank crews,” explained Ground Combat Systems Division Deputy Director Jami Davis.

Starting in August, the Army began bringing in small groups of tank operators to test out the new system, mostly using a new virtual reality setup that replicates the ATLAS display and controller. By gathering soldier feedback early, the Army hopes that they can improve the system quickly and make it ready for fielding that much faster. Already, the Army has brought in 40 soldiers. More soldier touchpoints and a live fire demonstration are anticipated to help the Army mature its product.

In some ways, ATLAS replicates the AI-capabilities demonstrated at Project Convergence in miniature. Project Convergence is the Army's new campaign of learning, designed to integrate new sensor, AI and network capabilities to transform the battlefield. In September, the Army hauled many of its most advanced cutting edge technologies to the desert at Yuma Proving Ground, then tried to connect them in new ways. In short, at Project Convergence the Army tried to create an environment where it could connect any sensor to the best shooter.

The Army demonstrated two types of AI at Project Convergence. First were the automatic target recognition AIs. These machine learning algorithms processed the massive amount of data picked up by the Army's sensors to detect and identify threats on the battlefield, producing targeting data for weapon systems to utilize.

The second type of AI was used for fire control, and is represented by FIRES Synchronization to Optimize Responses in Multi-Domain Operations, or FIRESTORM. Taking in the targeting data from the other AI systems, FIRESTORM automatically looks at the weapons at the Army's disposal and recommends the best one to respond to any given threat.

While ATLAS does not yet have the networking components that tied Project Convergence together across domains, it essentially performs those two tasks: It's AI automatically detects threats and recommends the best response to the human operators. Although the full ATLAS system wasn't hauled out to Project Convergence this year, the Army was able to bring out the virtual prototyping setup to Yuma Proving Ground, and there is hope that ATLAS itself could be involved next year.

To be clear: ATLAS is not meant to replace tank crews. It's meant to make their jobs easier, and in the process, much faster. Even if ATLAS is widely adopted, crews will still need to be trained for manual operations in case the system breaks down. And they'll still need to rely on their training to verify the algorithm's recommendations.

“We can assist the soldier and reduce the number of manual tasks that they have to do while still retaining the soldiers' ability to always override the system, to always make the final decision of whether or not the target is a threat, whether or not the firing solution is correct, and that they can make that decision to pull the trigger and engage targets,” explained Deaver.

https://www.c4isrnet.com/artificial-intelligence/2020/10/29/how-the-army-plans-to-revolutionize-tanks-with-artificial-intelligence/

Sur le même sujet

  • Stackley: Combined L3Harris Technology Will Compete to Build New Navy Distributed Battle Networks

    5 août 2020 | International, Naval

    Stackley: Combined L3Harris Technology Will Compete to Build New Navy Distributed Battle Networks

    By: Megan Eckstein August 4, 2020 3:25 PM A year after L3 and Harris merged into a single $18-billion defense company, the corporation is finding its formerly siloed components can come together to meet some of the Navy's and joint force's most complex needs. Sean Stackley, president of the Integrated Mission Systems segment for L3Harris Technologies, told USNI News in an interview that L3 and Harris each had important pieces of the puzzle to help the Navy achieve its distributed maritime operations concept. But Stackley, who previously served as the assistant secretary of the Navy for research, development and acquisition from 2008 to 2017 and as the acting secretary of the Navy from January to August 2017, said the key to DMO is not just fielding new platforms and tools but rather managing how information flows throughout the network, he said. Under the Navy's DMO vision, rather than deploying concentrated strike groups to a few places around the globe, the Navy would have many dispersed ships and planes that could share data to create a combined picture of the battlespace. He described the future fight as a combination of aircraft, ships, submarines and ground vehicles – manned and unmanned – all with sensors and communications devices, feeding data into a battle management system. The challenge will be the ordnance-to-target ratio and picking out the right targets to control the fight. Before the fight starts, the U.S. needs to ensure it has control of the EM spectrum so that network of platforms can communicate, sense and target. “It's really about linking sensors, providing assured communications, having the ability to disrupt the enemy's communications in their operating picture. It's everything from electronic support to electronic attack. ... That is a tremendous challenge because you have to work across the services, work across the platforms, you have to work across industry, you have to work across systems. So there's not one contract that's going to go out for DMO; it's going to be incremental. It's going to be an incremental approach to building this capability over time, over systems. And frankly the Air Force and the Navy are taking different approaches. I think there are some best practices across the services that they'll benefit by as each of these get more mature,” he explained, saying those were his personal views and not the company's. “I'm frankly studying the way the Air Force is approaching ABMS [Advanced Battle Management System], and I see a lot of strengths to their approach. There's a lot of parallel activities to the way they're contracting ABMS that should allow, if we do it right, should allow the incremental steps that need to be taken to be done in parallel as opposed to one at a time in a series. And I'm frankly also spending time with the Navy trying to link up the Navy's approach to DMO with the Air Force's approach to ABMS, to at least study – the services should be studying each other's approaches – and best practices should emerge, because otherwise we won't get there, it will take too long.” For example, he said, the Navy is preparing to contract for a ship-based signals intelligence program called Spectral. It also has an upcoming competition for a Spear program for electro-optical/infrared targeting. Under DMO, Stackley said, those two could be approached in parallel to ensure the whole network has access to the data they produce, instead of pursuing them separately and waiting for someone down the line to integrate the systems into a larger network. “Traditional (acquisition) says you do the standalone upgrades; inside of DMO, you're constantly looking at the total framework architecture, how do these capabilities integrate” on the front end “so that on the back end you are, in fact, building a distributed maritime operational capability,” he said. Stackley said the company is positioned to adapt to the changing requirements of DMO. “We are on the ocean floor, and we operate from the depths of the sea to the depths of space. We are in every domain. We operate across the entire kill chain, from sensing, communications, tracking, targeting, right down to putting ordnance on target. We operate across the kill chain and across the entire electromagnetic (EM) spectrum. In the acoustic realm, we operate below 10 hertz, and then you move into the [radio frequency] and in the RF end of the EM spectrum we're operating above 50 gigahertz. So we dominate – I would say spectrum superiority is one of our strengths. And we do this to provide capabilities, solutions, for national security, ours and our allies.” The company's advantage is based on “two companies a couple of years ago that had a large number of stand-alone capabilities seeing a match in terms of our separate capabilities, and also seeing the power that comes through integration of these capabilities, understanding where the customer is going in terms of the future fight where that EM spectrum, that spectrum superiority, is so critical. Whether you're talking about the Navy's strategy, the Navy's vision for distributed maritime operations, or the Air Force's advanced battle management system, it is the same capability the services are looking for, which is to have the advanced sensors at the forward edge, have the information that they collect communicated back through secure data links to platforms, have that information integrated into a common picture so that we can control the spectrum, we can ensure our communications, we can disrupt [adversaries'] communications, and we can pull the information from our sensors and get it to where it's most needed so that when the time comes we can put ordnance on target rapidly and reliably,” Stackley said. The two companies had different tools in their portfolios prior to the merger that contribute to this new ability to network together tools for fighting in the EM spectrum. For example, “Harris focuses on tactical communications, electronic warfare, space payloads and supports FAA air traffic control modernization. L3's portfolio is a bit more diverse and includes electronic components, aircraft modernization, flight simulation, UAS/UUVs, airport security and C4ISR components and subsystems,” Defense News quoted Byron Callan, an analyst for Capital Alpha Partners, as writing in a note to investors ahead of the merger. In the interview, Stackley used undersea warfare as an example of where L3 and Harris have been to provide the Navy options to support DMO. On the seabed, the company leveraged each of the halves' legacy systems to create an underwater acoustic system that won a prime contract with the Navy – something neither L3 nor Harris could have done before the merger. “Within the first year, we're offering integrated solutions to the customer that prior to the merger we would never have seen and would never have found together,” Stackley said. The combined portfolio also includes experience with unmanned underwater vessels. L3Harris is competing for the Medium UUV program that will replace separate medium UUV systems for the explosive ordnance disposal and the submarine communities. Stackley said the company had an already-existing, highly modular design that allowed it to work with Navy labs to integrate and operate advanced payloads at sea while the Navy was developing its specifications for the MUUV program. The company's UUV experience, Stackley said, coupled with underwater acoustic systems and above-water communications capabilities that reside within L3Harris, means it can offer a package that allows the Navy to receive real-time or near-real-time updates from this UUV. The company also recently won a contract with the Navy to design and build at least one Medium Unmanned Surface Vehicle (MUSV), with options for more vehicles. Stackley said L3Harris had extensive experience with USVs, including through the Overlord large USV demonstrator program run by the Pentagon's Strategic Capabilities Office. For its MUSV offering, the company is partnering with Gibbs and Cox, which also participated in the Overlord program. Through its in-water testing, L3Harris has learned about autonomy software, vehicle reliability, and command and control. Stackley said the company, outside of the MUSV program, wants to take its USV a step further and demonstrate to the Navy another option for combining several legacy L3 and Harris technologies. The company builds the signals intelligence system on the Air Force's RC-135 surveillance aircraft. That system had been stovepiped in the company's aircraft systems division before, but Stackley said L3Harris plans to use that as the basis for the upcoming Spectral competition, which will be a ship-based SIGINT tool. L3Harris will adapt that system for integration on a medium USV, he said, thereby demonstrating “a sensing capability, where you start with a reliable unmanned surface vessel that has endurance on station, more so than an aircraft; you give it a sensor package that [meets Navy and Joint Force needs]; and then you add to that the data links that L3Harris provides and the secure communications that we provide, so that now you've got a node on the network that's passing critical information to the operating force from an unmanned vessel.” He made clear that the SIGINT package on the USV is not part of the Navy's current MUSV program but that L3Harris would pitch the capability to the service. https://news.usni.org/2020/08/04/stackley-combined-l3harris-technology-will-compete-to-build-new-navy-distributed-battle-networks

  • Epirus wins $66M Army contract for drone-frying Leonidas microwave kit

    23 janvier 2023 | International, Aérospatial, C4ISR

    Epirus wins $66M Army contract for drone-frying Leonidas microwave kit

    High-power microwave systems use bursts of energy to disrupt, degrade or destroy distant electronics.

  • NATO Secretary General launches his Annual Report for 2023

    14 mars 2024 | International, Sécurité

    NATO Secretary General launches his Annual Report for 2023

    NATO Secretary General Jens Stoltenberg launched his annual report for 2023 on Thursday (14 March 2024), which covers all aspects of the Alliance’s work over the past year.

Toutes les nouvelles