Synth Wave: training in the clouds

Pilots can train together as a distributed network around the world in a synthetic environment. Credit: Nato Support and Procurement Agency / NFTE

As technology develops, more and more of training is taking place in simulated and synthetic environments, where personnel can improve on skills repeatedly for optimal performance, largely risk-free.

For that reason, annual US investments in this market are forecast to exceed the $26bn mark in 2028 according to GlobalData’s figures.

In particular, air forces are leaning on simulation for flying training. In 2021, the UK Royal Air Force (RAF) said it conducted around 50% of its training in a simulated environment and aimed to increase this to 80% by 2040. Currently, trainees spend more than 60 hours in synthetic training systems in preparation for real-world flight.

There is a case for this globally, too. Nato Flight Training Europe (NFTE) reached full operating capability in mid-November 2024.

This initiative – boasting 13 members, with more nations expected to join in 2025 – coordinate its shared facilities for military flying training across the continent for fast jet, transport, helicopter and remotely piloted air vehicles. Clearly, capacity is a growing issue that is beginning to be addressed.

“We now have the framework and processes required to maximise the effectiveness and minimise the administrative burden when sharing training capacity between nations,” assured Group Captain Ryan Morris, RAF, NFTE Steering Board Chair.

However, military air training is becoming more complex in this decade as aerial warfare evolves, especially with the rise of autonomy. Therefore, more thought is needed to gauge the ideal balance between real-world and simulated training; the vital role that data plays; and how to accommodate the new hybrid force structure for sixth-generation tactical air combat. 

“Physical, live training is the most valuable and efficient, but it is always the hardest and most complex to do, logistically and financially,” suggested Tristan Sauer, GlobalData defence analyst. 

The UK Ministry of Defence (MoD) is confronting these enduring problems as it tries to manage the training of personnel across the armed forces.

In the RAF, trainee pilots are facing extended times in ‘holdover’, the period when one training course has finished and another has yet to start. In conjunction, the government lacks the capacity for training at the point of demand – new, mobile firing range sites are being procured to meet the growing demand.

It appears that alternative training tools are used, in this case, to plug holes in the defence estate rather than to optimise training as an end in itself.

The former Chief of the Air Staff, Air Chief Marshal Sir Mike Wigston, explained to the previous Defence Select Committee in February 2023 that the service faced a major training backlog after the Covid-19 crisis.  

However, it should be noted that this air-launched missile – more accurately considered an aeroballistic missile as it is not strictly ‘hypersonic’, lacking manoeuvrability at excessive speeds – is believed to be one of the least sophisticated weapons in Russia’s hypersonic arsenal.

Speaking to Global Defence Technology, William Freer, research fellow in national security, Council of Geostrategy, provided his assessment of the fledgling Russian Kinzhal:

“Initially it seemed that the Russians had the right idea in targeting Patriot batteries with their Kinzhals… However, it seems the Patriots were able to deal effectively with this threat, so the Russians no longer use them in this manner as far as I am aware,” Freer said.

Even the best laid plans do not survive the battlefield. Therefore, the Kinzhal was repurposed to enhance conventional missile salvos after their inability to penetrate Western air defence technologies on their own, despite Russian President Vladmir Putin’s bombastic praise of the Kinzhal since it was first produced in 2018. 

Subsequently, Wigston made the case for synthetics as a shining solution to these lingering problems. He even left open the possibility that synthetics could replace the Hawk T2 training aircraft, which approaches the end of its life cycle in 2040.

This prospect resurfaced once again, in January 2024, when the Committee grilled British industry over the potential future training gap. At the time, Simon Barnes, group managing director air, BAE Systems, was similarly open-ended, suggesting that there will be a “transformation” in flight training in terms of the “balance between live and synthetic”.

However, simulated training offers far more than the ability to compensate the armed forces with a cheap, on-demand alternative to real-world training. While it is true that military training in the synthetic environment offers the ability to improve minute aspects of flying, this should not disrupt efforts to overcome the government’s inability to manage the time, cost, and logistics of real-world flying and training requirements.

The value of simulation training lies, also, in the data and analytics captured from repeated use cases. “The more data you have, the better your insights are going to be,” Niall Campion, managing director, VRAI, told Global Defence Technology.

A start-up that arms instructors and trainees with actionable insight based on extensive simulation training datasets through its flagship product Heat, VRAI has worked with the RAF since 2020. 

The company has captured more than 900 million data points from 39 active-duty pilots, and this has coloured its growing picture of British airmanship. However, this is not enough to generate insight, Campion stated.

This is why air forces must be able to draw on decades of simulation training data to shape the optimal performance of trainee pilots. “In defence, the biggest bottleneck at the moment is access to these data sets,” he added.

“The analogy I like to use is ChatGPT or the Large Language Models that are being used at the moment. They’re trained on billions of pages of text. In fact, some would say they’re trained on all the Internet. There is no equivalent dataset in defence, for obvious reasons (security, among others), but the main reason is that no one is capturing it,” Campion explained.

The RAF has been running simulations for the last 30 years. While it previously used stand-alone simulators for each aircraft type, the service now creates networks of distributed synthetic multi-domain integration training through the Gladiator capability and the Air Battlespace Training Centre.  

These assets enable realistic large force, multinational exercises in the synthetic environment. 

Yet in all that time, the RAF is left with no data from what would have been enough to inform any actionable insight today: “The data is gone, you have lost 30 years of insight,” said Campion.

This is ironic considering the MoD has, since 2020-21, identified data as a “strategic asset”. Since then, though, the Integrated Review Refresh, published in May 2023, also touched on the possibility supporting data-sharing infrastructure and removing barriers to global data access and use.

Should the new Labour government favour this policy in the Strategic Defence Review in the spring of 2025, it could help the UK interoperate with allies through initiatives such as the NFTE, or perhaps AUKUS pillar two. 

Looking forward, it will be difficult to simulate contemporary advancements in aerial warfare with the rise of sixth generation air combat. Right now, it is still uncertain as to what this will look like; it is still largely undefined.

In general, however, major air forces are investing heavily – perhaps too heavily in some cases, such as the UK’s Global Combat Air Programme – in crewed-uncrewed teaming.

At the heart of the future network, a sixth-generation crewed fighter jet will operate alongside autonomous, uncrewed aerial vehicles (UAVs), that offer attritable mass at the tactical level. 

When asked how users will replicate the concept on the simulator, Dr Al Allsop, Combat Air and Synthetics SME at British defence training supplier Inzpire, stated “we do not know yet.”

The crewed-uncrewed challenge may be addressed “with the [crewed] pilots [and] UAV teams conducting their training initially in the simulator. Then, once they move to live flight, the UAVs remain in the simulator with their positions, decisions and output of effects relayed to the live pilots by means of the [live, virtual and constructive] network,” Allsop said.

This intricate process demonstrates the type of emerging technologies, such as artificial intelligence (AI), that the military have yet to truly grasp and implement fully. The complexity of crewed-uncrewed teaming will put more pressure on simulators, and replicating various autonomous systems will require a lot of computing power.

“If we assume the UAVs will have advanced AI in them rather than just being used as a relay for sensors sent into dangerous environments, then the answer is yes,” said Allsop, regarding the potential increase in power requirements for simulators.

“Simplistic decision algorithms (think logic gates) could be housed within the same programme, more probabilistic based or ‘trained’ AI would likely require separate instances for each UAV represented. Each of those requiring its own representation of the input of its sensors,” Allsop detailed.

While this force multiplier model will be difficult to replicate, it is not beyond possibility. MBDA UK, one of Europe’s complex weapons specialists, opened its fully operable Digital Battlespace Facility last year.

The site, in Stevenage, leverages digital twin technology to allow users to design, troubleshoot, and enhance a mix of different weapon systems together in the company’s portfolio with different platforms in a virtual, risk-free environment.

The future weapon systems sector will require a digital backbone to showcase the interoperability and interactions of different systems. 

Australia could be one of the main beneficiaries of this dramatic increase in demand, where private companies and local governments alike are eager to expand the country’s nascent rare earths production. In 2021, Australia produced the fourth-most rare earths in the world. It’s total annual production of 19,958 tonnes remains significantly less than the mammoth 152,407 tonnes produced by China, but a dramatic improvement over the 1,995 tonnes produced domestically in 2011.

The dominance of China in the rare earths space has also encouraged other countries, notably the US, to look further afield for rare earth deposits to diversify their supply of the increasingly vital minerals. With the US eager to ringfence rare earth production within its allies as part of the Inflation Reduction Act, including potentially allowing the Department of Defense to invest in Australian rare earths, there could be an unexpected windfall for Australian rare earths producers.