[Popular Science of Chinese Military Technology] How can “smart” AI improve the accuracy of intelligent aiming in combat?

As we enter 2023, AI technology continues to develop rapidly, and people are surprised to find that AI can be so "smart". As the most cutting-edge military technology field, can AI technology help solve multiple "pain points" including aiming? Can a sharpshooter who needs long-term hard training become a "100% accurate" person with the help of AI?

Under the wave of AI technology, not only is auxiliary AI aiming technology booming, but attempts at fully autonomous AI aiming and shooting are also underway. In terms of auxiliary aiming, many countries are using AI technology to solve the "pain points" of their platforms one by one based on existing mature technology platforms, and artificial intelligence is being installed on a variety of weapon platforms for testing.

The "Atlas" artificial intelligence assisted aiming system is installed on the top of the tank (Source: Huanqiu.com)

Taking the famous American "Abrams" tank as an example, military technology experts recently installed an artificial intelligence system on the "Abrams X" main battle tank. In May 2023, the US military released a set of photos on its Defense Visual Information Distribution Service (DVIDS) website, showing an "Abrams X" main battle tank equipped with the "Atlas" artificial intelligence-assisted aiming system during an exercise in November 2022. The "Atlas" software system is an artificial intelligence machine learning algorithm that continuously performs object detection and image classification by receiving battlefield data sent by the optoelectronic turret and sensors, automatically selecting targets, analyzing target types, speeds and distances, finding targets that pose the greatest threat to the tank, and then sending these images to the touch screen inside the turret.

With the completion of a series of AI-assisted aiming, all the gunner has to do is select the target with a touch of his finger, and the tank will automatically rotate the turret and aim the gun at the weak point of the selected target at all times. At the same time, AI will automatically recommend suitable ammunition. The gunner only needs to approve the AI's suggestion, and AI will aim on its own. Calculate the shooting elements, and then open fire. Although humans are involved in the whole process, most of the work that needs to be undertaken has been shared by AI technology. In addition to army equipment, the air force is also an important area for AI technology intervention. For example, the MQ-28A "Ghost Bat" being tested in Australia has an artificial intelligence system that can control its flight route and combat mode by itself, and provide assistance in electronic countermeasures, surveillance, and strike targets to increase hit accuracy and enhance surveillance and electronic countermeasures.

MQ-28A drone in test flight (Photo source: NetEase)

In addition to assisting aiming, another important direction for AI technology in the future is to fully take over human "aiming", that is, to successfully achieve identification, aiming, and attacking without human intervention. If AI wants to achieve this full-process operation, the first step it needs to take is identification, using high-precision sensors such as radar and cameras to identify the target to be attacked, and at the same time calculate its movement state and trajectory, and then track and lock the aiming through image technology. Because there is no human intervention in this process, AI is required to make a comprehensive judgment on the target's speed, direction, distance, posture, and even deceptive information such as enemy and friend.

In June 2021, the U.S. Navy publicly announced that the MQ-25A Stingray, equipped with an autonomous unmanned system, completed a test of aerial refueling of a manned F/A-18F. (Photo source: China Aviation News)

Take an earlier famous report as an example. The Hong Kong South China Morning Post website published a report titled "Will they become the new first-class pilots in future aerial dogfights?" The artificial intelligence "pilot" created by the US defense contractor Herron Systems swept a US Air National Guard pilot and a graduate of the US Air Force Ordnance School with 2,000 hours of F-16 fighter pilot experience with a score of 5 to 0. But then military analysts said that this result does not mean that artificial intelligence can defeat or even replace humans in real combat. Existing artificial intelligence technology is still far from actual combat application, because its response is based on big data technology and cloud computing, and the scene is simulated through algorithm optimization. The amount of data and computing power will directly determine the actual effect. This also seems to indicate that there is still a long way to go for fully autonomous AI decision-making.

In 2017, many AI technology company leaders called for a ban on the use of AI robots to eliminate manpower in war. In 2018, Google announced that it would terminate its cooperation with the US military in using artificial intelligence to analyze drone videos, and promised not to use AI technology for weapons development. There is no doubt that at this stage, AI aiming technology is still unable to make a complete series of "observation, adjustment, decision-making, and action" instructions, but with larger-scale data learning and stronger computing power, AI-based aiming technology will become more and more reliable.

Expert profile: Wang Teng, a popular science expert on military equipment, holds a master's degree from the Army Armored Forces Academy. He has worked in frontline military news at CCTV and People's Daily, and has participated in the planning and production of many military science and technology topics. He has been conducting popular science research on domestic and foreign high-tech equipment and main combat equipment of various arms for a long time. He has published articles on platforms such as People's Daily Overseas Edition, Legal Daily, and Science Popularization China, and won the Advanced Individual Award issued by the "National Informatization Leading Group".

Produced by: Popular Science of Chinese Military Technology

Author: Wang Teng (senior military science writer)

Scientific review: Wang Mingzhi (military expert)

Planning: Jin He

Producer: Guangming Online Science Department