When the words military and AI are used together in a sentence, often the first thing that comes to mind are drones or sophisticated fighter jets. Innovation out of the U.S. Army may change that immediate association, however, as it recently announced it is equipping tanks with artificial intelligence. The new technology, known as the Advanced Targeting and Lethality Aided System, or ATLAS, uses machine learning algorithms and advanced sensors.
“Currently, tank crews use a very manual process to detect, identify and engage targets,” an army Master Gunner told the military website C4ISRNet this month. “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.”
AI can automate that process and make it much more efficient, enabling targets to be found and fired upon much more expediently. The ATLAS system is ‘device agnostic’ meaning it can be installed in many different tanks. An optical sensor is installed on top of the vehicle and sends information on a potential threat through a machine learning algorithm. Once a threat is identified, images of it are sent to a touchscreen device inside the tank, viewable by Army personnel. It is then up to the military operator to carry through with the AI recommendation and manually press fire on the target. The operator can also decide to change the AI recommended settings, or to disengage the suggested action.
“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,” says the C4ISRNet writer who attended a demonstration of ATLAS at an Army facility in Maryland. “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.”
ATLAS was developed by the Army C5ISR Center – dedicated to research and development relating to Command, Control, Communications, Computers, Cyber, Intelligence, Surveillance, and Reconnaissance – together with the Military Armaments Center. After the system was developed, 40 Army tank operators were asked to provide feedback which was incorporated into refining ATLAS. An exact figure on how much quicker ATLAS can aid in destroying targets than humans carrying out the same process is unknown, though it has been suggested that it may be as much as three times faster. That time frame could be reduced further, if the human operator did not have to give a manual command to fire. But as Breaking Defense magazine elaborates, that would be against military rules.
“ATLAS could be faster still if the Army were willing to take the human gunner out of the loop and allow the AI to fire the gun,” Sydney J. Freedberg Junior writes in Breaking Defence. “That won’t happen, unless U.S. policy changes. The American military – unlike Russia and China – is profoundly wary of what autonomous weapons could do without human supervision. The US sees its highly trained troops as an asset that AI should empower, not an obstacle for AI to bypass.”
Dawn Deaver, the project leader for ATLAS at C5ISR agrees. “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,” Deaver says. She is focused on increasing the efficiency of Army personnel, not replacing them. “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.”
The outcome is a fine example of how AI and humankind can work together to improve processes, rather than one assuming the job of the other. This symbiosis paves a path for how we can leverage the best from AI without being overtaken by it.
Artificial Intelligence is being implemented in industries all over the world and is a central theme of the research undertaken at UCIPT. Our work in the HOPE study is using data to assess and shift behavioral outcomes among HIV and other populations.