Imagine a retreating army being chased by dozens of small drones while fleeing, at first glance indistinguishable from amateur drones.
These devices are equipped with cameras that scan the area and processors that autonomously determine targets.
The drone starts destroying trucks and individual soldiers, exploding on contact with objects.
This is not science fiction about future wars, but events that have already occurred in the spring of 2020. Soldiers loyal to Khalifa Haftar then retreated from Turkish-backed Libyan government forces.
According to UN officials, the drones were “stalking” the fleeing militia.
Then, the United Nations recorded the first case in history of a drone killing a person without direct orders from the operator.
This is just one example of the use of autonomous systems on the battlefield, and there are many other applications for a variety of missions.
ForkLog took a look at the capabilities of AI in military affairs and how militaries of different countries are using the technology in practice.
- The Department of Defense uses algorithms to recruit soldiers, make strategic decisions, identify threats, and more.
- The Ukrainian Armed Forces will use biometrics, sapper robots, tools to search for suspicious persons and tools to analyze satellite imagery.
- The Israeli military uses artificial intelligence turrets, drone swarms and target elimination systems.
- The US is testing AR glasses for the military, unmanned helicopters and “killer robots”, the Netherlands is testing autonomous armored personnel carriers, and China is testing algorithms to predict the course of missiles.
- A lack of training data and soldiers’ mistrust of the technology hinder the militarization of AI.
Artificial Intelligence Algorithms in the Military Field
Artificial intelligence and machine learning algorithms play a key role in military operations.
The military started using the technology even before it became commonplace in civilian life. All the while, engineers have been actively testing many algorithms, developing them, improving them and teaching them to perform increasingly complex tasks.
Today, the military can use AI to:
- recruit soldiers;
- prepare the army for the mission;
- making strategic decisions;
- data processing and research;
- Object detection and recognition;
- eliminate the target;
- threat monitoring;
- transport of persons and goods;
- assist doctors;
- cyber security.
Over the years, Defense Department representatives have used a variety of methods to recruit soldiers, from booths at job fairs to posters and TV ads.
A poster calling for service in the U.S. Army during World War I. Data: Library of Congress.
Artificial intelligence is making it easier to find qualified candidates ready to serve in the military by quickly crunching large amounts of data and automating and streamlining aspects of the recruiting process.
Artificial intelligence can help train soldiers in various skills. In February, for example, technology company Northrop Grumman awarded DARPA a contract to develop an artificial intelligence assistant to train Black Hawk helicopter pilots.
The system is expected to be embedded in an AR headset, and through voice and graphic indicators, will help pilots learn new missions, reduce errors and speed up missions.
A visualization of the pilot’s AI assistant working. Data: Northrop Grumman.
The U.S. Army also uses combat training simulation software. The software allows soldiers to complete tasks in VR and acquire skills they can apply in real life.
AI systems that can collect and process data from multiple sources help the military make strategic decisions. Algorithms can learn different scenarios quickly and efficiently, recommending the best course of action even in critical situations.
Army analysts use neural networks to search for the right information in news articles and social networks, filtering, sorting and identifying only reliable information. This allows you to optimize and speed up the research process.
In combat, soldiers use computer vision systems to detect and identify objects. Algorithms can identify opponents, find their weaknesses and predict behavior. In addition, some artificial intelligence-based devices are capable of destroying targets after they are identified, at the command of the operator.
The battlefield is an extremely dangerous place. Drones, robots and other threat monitoring systems enable the military to quickly assess risk and receive guidance on how to minimize it.
The transportation of ammunition, weapons and cargo is critical to the success of military operations. Armored robotic vehicles without human intervention can transport objects, determine the best route under current conditions, and collect necessary information while moving.
If a soldier is injured, special artificial intelligence systems are able to recommend treatment. These algorithms can access hospital databases to help doctors make decisions in stressful situations.
Information systems storing information about soldiers, dignitaries, missions and technology are regularly targeted by cybercriminals. Algorithmic programs are able to detect threats in advance and form strategies to protect important data.
How is Ukraine using artificial intelligence in war?
In the past few years, more and more countries have begun to develop various combat equipment and artificial intelligence systems. Among them: “killer robots,” AI turrets, unmanned helicopters, AR headsets for soldiers, “smart” rifles and tools to identify war criminals.
Some AI solutions have been successfully tested by the armed forces of different countries. One of the most obvious examples is Ukraine.
On February 24, 2022, Russian troops invaded the country, starting an all-out war that is still ongoing. It has become Europe’s biggest land conflict in nearly 80 years.
Despite Russia’s numerical and firepower superiority, Ukrainian forces are driving the enemy from the field. Technology has played a major role in this.
On February 25, AI startup Reface announced that it had created an algorithm to identify Russian troops from satellite imagery. To gather training data, the developers asked Ukrainian citizens to send in new messages.
The company also sent 13 million anti-war messages to Russian users through its app, bypassing Kremlin censorship and calling for support for sanctions against the aggressor country.
In March, the Ukrainian Ministry of Defense began using Clearview AI’s facial recognition system.
The company has allowed its technology to be used to identify the Russian military, fight disinformation, identify the dead and screen people at checkpoints.
Ukraine’s Minister of Digital Transformation Mykhailo Fedorov noted that the government will also use other facial recognition systems. According to him, these algorithms help to identify the Russian military involved in the attack on civilians in the Kyiv area.
In April, YouControl, a developer of open data-based services, and Artellence, an artificial intelligence company, launched the TiKhto app with the support of the SBU, which allows you to quickly identify suspicious persons.
The system can determine whether a passport photo is genuine, invalid or considered lost. It will also show whether the person being checked is on the country’s most wanted list, is on the list of terrorists or peacemaker bases under sanctions imposed by the National Security and Defense Council.
The operating principle of the application “quiet”. Data: You control.
That same month, first deputy interior minister Yevgeny Yenin said AI technology helped police detain around 200 members of illegal armed groups.
In June, the U.S. Army agreed to donate Boston Dynamics’ Spot robot dog to Ukraine to help clear mortar shells and cluster munitions from former Russian-held areas. The device uses a robotic arm capable of dragging unexploded mines into special pits for neutralization.
In August, Novaya Poshta and the National Emergency Service announced the development of a robot engineer. The company plans to produce at least 40 devices for finding and clearing landmines in bodies of water and on land.
In October, Deep State UA, developer of an interactive online map of Ukrainian hostilities, announced a partnership with creators of Griselda, an automated neural data processing system.
Israel and its combat artificial intelligence
In the spring of 2021, soldiers of the Israel Defense Forces took the lead in using AI-controlled drone arrays in real combat conditions. They deployed a drone squadron to locate, identify and attack Hamas militants in the Gaza Strip.
Then the Israeli military also used artificial intelligence and supercomputers. The algorithm provides recommendations soldiers use to identify targets and engage them.
Another machine learning model warns troops of possible enemy attacks.
That fall, news emerged that Israeli intelligence had killed Mohsen Fakhrizadeh, a prominent Iranian nuclear scientist, with a remote-controlled machine gun.
In 2020, secret agents installed an automatic machine gun on a pickup truck along the scientist’s route near Tehran. As Fakhrizadeh’s car approached, the remote operator opened fire, issuing commands via satellite. Artificial intelligence helps compensate for signal delays and feedback.
Iranian researchers said the operatives may have used facial recognition technology to locate the targets.
In September 2022, the Israeli company Elbit Systems launched the ARCAS artificial intelligence system to make it easier for the military to shoot.
The complex includes a sight attachment and a helmet-mounted eyepiece, allowing soldiers to follow information about the battle in real time.
A rifle equipped with the ARCAS system. Data: Elbit Systems.
That same month, Israeli authorities installed remote-controlled weapons at Hebron checkpoints that could fire stun grenades, tear gas and sponge bullets. The turret uses an artificial intelligence system to track and lock on to the target.
According to army representatives, the complex does not include remote control of live ammunition.
The system was installed at the checkpoint in Hebron. Data: Haaretz.
In October, the military deployed an autonomous target-tracking turret on a watchtower aimed at the Al-Arub refugee camp in the West Bank.
According to the developers of the complex, the purpose of the system is to protect soldiers and civilians by increasing the accuracy of hits.
In November, Elbit Systems unveiled Lanius, a lethal autonomous kamikaze drone controlled by artificial intelligence.
The maneuverable, fast drone is equipped with numerous sensors, cameras, and an Nvidia Jetson system for reconnaissance, mapping, and object classification.
The company stresses that the quadcopter will not be able to eliminate the target at its own discretion. There is always someone involved in the process.
In 2023, the Israeli Ministry of Defense plans to begin testing unmanned M-RCV tanks equipped with autonomous 30mm gun mounts and reconnaissance drones.
Military AI Developments in the U.S. and Other Countries
In addition to Ukraine and Israel, the United States and other countries are also actively developing the field of combat AI.
One of the interesting American projects is an augmented reality headset for soldiers. Her story is full of ups and downs.
In spring 2021, Microsoft won a tender to provide the U.S. Army with a prototype of an Integrated Visual Augmentation System (IVAS) based on a HoloLens AR headset. The potential value of the contract is close to $22 billion over 10 years and provides a supply of 120,000 units.
The headset integrates high-resolution nighttime, thermal and soldier wearable sensors. The information is displayed on the helmet visor.
The system uses AR technology and artificial intelligence based on Microsoft Azure cloud services to create a realistic mixed reality learning environment.
Prototype of an integrated visual enhancement system based on a HoloLens AR headset. Data: Microsoft.
In October 2021, the U.S. military decided to postpone the deployment of IVAS until May next year. They said they took that time to prepare the system for the intensive testing required for full production.
In January 2022, the Pentagon said the HoloLens-based headset was a “promising project” but not yet ready to be tested in combat conditions. In May, the U.S. Department of Defense announced a reduction in the cost of the IVAS procurement contract.
In September alone, the US military received the first 5,000 prototypes of the integrated visual enhancement system. But a month later, the media reported cases of illness among soldiers using IVAS.
The military admitted there was a problem with the headset and announced it had been fixed.
As for other U.S. military projects, in February 2022, DARPA tested the Black Hawk UH-60A unmanned combat helicopter. The car flew itself for 30 minutes over its Fort Campbell base in Kentucky.
The device developed by Sikorsky is equipped with the experimental ALIAS AI system. According to the developers, it’s a “customizable, detachable flight automation suite” that combines hardware and software to reduce crew workload.
In November, the agency retested an unmanned helicopter, training it to transport people and cargo.
That same month, Lockheed Martin and Microsoft entered into a strategic partnership to develop next-generation technologies for the U.S. Department of Defense. The agreement covers four key areas, including artificial intelligence, modeling and simulation.
Several U.S. companies, including Ghost Robotics, are actively building robots that can carry rifles and other weapons.
Quadruped robot Q-UGV from Ghost Robotics, equipped with a special-purpose automatic rifle. Data: Jian International.
Other countries also have interesting examples of military AI development.
The Netherlands is the first NATO member to conduct tests of an armed unmanned ground vehicle (UAV) in Lithuania.
The devices were created by Estonian defense company Milrem Robotics. BNTS is the THeMIS tracked hybrid modular infantry system capable of using a variety of weapons, including machine guns.
In the UAE, Emirates Defense Industries is developing a compact launcher with a smart self-destructing drone.
The system is the base for launching the Hunter 2-S artificial intelligence drone. These devices can fly in formation and “communicate” with each other to carry out complex joint attacks on their own.
British researchers will create a system of VR headsets and medical robots to provide first aid to wounded soldiers on the battlefield.
Frontline doctors will be able to don virtual reality headsets and operate autonomous devices to collect smears and blood samples, measure temperature and blood pressure, the scientists said.
In China, researchers developed an artificial intelligence algorithm to predict the course of a hypersonic missile. The system does not require huge computing power and produces results within 15 seconds.
The technology was able to determine the path of the missile moving five times faster than the speed of sound, calculate the potential trajectory to hit it, and launch a counterattack with a three-minute lead.
Why is the militarization of AI slow?
War zones are arguably one of the most technically challenging areas to deploy AI systems. This is due to the lack of a large amount of relevant training data, as well as soldiers’ mistrust of the technology.
Although many countries are actively promoting the idea of using artificial intelligence, it is difficult for them to move from concept to implementation. That’s partly because the defense industry in most states is still dominated by large contractors with no experience developing smart software.
It’s also because the “clumsy” government review process moves too slowly compared to the dizzying pace of technological development. Military contracts can last for decades, but most startups only get started for a year or so.
Additionally, not all states have addressed ethical issues related to the use of artificial intelligence. Only a few countries and federations have issued relevant guidelines.
All of these rules encourage the military to use technology in a legal, responsible, trustworthy and traceable manner. They also aim to remove bias inherent in the algorithm.
However, advances have been made in many defense-related fields due to the advent of appropriate artificial intelligence systems.
Nations are now vying for leadership in combat AI. And, perhaps in the not-too-distant future, the criterion for assessing a nation’s military superiority will no longer be the size of the army, but the performance of the artificial intelligence algorithms employed by soldiers.
Subscribe to ForkLog news on Telegram: ForkLog AI – All news from the world of AI
Found an error in the text? Select it and press CTRL+ENTER
Information source: compiled from FORKLOG by 0x Information.The copyright belongs to the author Марина Глайборода, and shall not be reproduced without permission