AI and 6G will form the bedrock of China’s first hypersonic missile defence system

While China has repeatedly demonstrated its hypersonic offensive capabilities, by showing off its DF-17 “carrier killer” hypersonic missile system and by shouting about its 6G based hypersonic control systems and hypersonic drone swarm programs, the Chinese government has now decided that it’s high time to develop some systems that can defend against these superfast weapons that today can’t be countered by anyone, and which are so fast that the US has said the only way to defend against them is to create fully autonomous Artificial Intelligence (AI) systems that can see, track, and respond in milliseconds.

So it’s no surprise therefore that Chinese military researchers claim to have developed Artificial Intelligence (AI) technology that can predict the trajectory of a hypersonic glide missile as it approaches its target at speeds exceeding five times that of sound, South China Morning Post reported.

Today rockets are used to launch what’s known as hypersonic glide vehicles which then carry on to obliterate their targets after separating from the rocket and moving toward its target at a speed of at least Mach 5, with some already achieving speeds of over Mach 20.

The Chinese researchers, however, seem to be several steps ahead of their American counterparts. According to them, a Chinese AI powered air defense system can predict the potential kill trajectory of an incoming weapon and launch a swift counterattack within three minutes. And, bearing in mind that the typical missile stays within an 8 kilometer, 5 mile, target zone, which is quite small for a weapon that can travel that distance in under two seconds, this means the AI could have a decent chance of intercepting it – provided of course it’s accurate and has the means to do so.

“The world’s military powers are currently engaging in a fierce arms race around the development of hypersonic glide vehicles and weapons, bringing new and severe challenges to air and space safety,” said Zhang Junbiao, a computer scientist from the early warning intelligence department of the Air Force Early Warning Academy in Wuhan.

“Trajectory prediction is of great significance to combat intent assessment and aerospace defense interception,” wrote Zhang and his team in the Journal of Astronautics, a peer-reviewed publication run by the Chinese Society of Astronautics.

These Chinese claims come on the back of a dedicated effort in the US to build and test a hypersonic weapon system to match the capabilities of China and Russia, the latter of which recently used their own hypersonic weapons in Ukraine.

So far the US has carried out a successful test of an Air-launched Rapid Response Weapon (ARRW) hypersonic missile, which was just a few days after it had successfully tested its Hypersonic Air Breathing Weapon Concept (HAWC). However, despite the progress it could be a while before we see the US and its allies put hypersonic weapons into the field for real.

A hypersonic glide weapon, unlike a normal ballistic missile moves through the atmosphere like a stone skipping across water and by banking left and right which makes it more difficult to detect and intercept, according to SCMP so, a Mach 5 or higher speed, there is little time for an air defense system to respond which is why it’s widely claimed that current technology just can’t stop these systems. Zhang, however, believes “AI is capable of handling such unpredictable tasks and developing a defense against an unpredictable trajectory and incredibly high speed.”

The defending side normally has no idea about the mass, size, shape, aerodynamic control system, or purpose of hostile weapons, but by analyzing observed flight data, the AI may make a fairly accurate assumption. According to the researchers, every move a missile makes will give off some modest but useful signals about its design, capabilities, and mission, regardless of how advanced or fast it is. As a result, a machine learning system may learn from data collected during the early phases of a hypersonic flight and utilize that information to forecast the most likely trajectory during the flight’s later stages.

Zhang and his colleagues acknowledged that translating this idea into a working model was difficult. They claim that the raw data collected by an early warning system contains a lot of noise that might confuse AI and that too much data can also overload the computer.

To combat the problem of unwanted and overwhelming noise that could mislead the AI, the Chinese researchers devised a novel deep learning method that could automatically eliminate noise from observed signals. The algorithm similarly simulates the human brain’s activity by focusing just on the most recent, most important data to save calculation resources.

The new system can run on a laptop computer and produce a result in 15 seconds, according to the study, despite being more advanced than any previous AI for hypersonic trajectory prediction. Simulated experiments indicated that the system is still effective against a wide spectrum of weapons traveling at up to Mach 12 speeds.

Earlier in March, China’s Aerospace Defense industry had claimed that Beijing had made significant progress in building an AI system that could even design new hypersonic weapons autonomously.

As tensions continue to rise between the US and China in the Indo-Pacific region with an aggressive Beijing practicing dangerous military manoeuvres policymakers and military are alive to the challenges posed by the Chinese hypersonic missiles, and as you can see the race to “be superior” is on like Donkey Kong as they say.