USAF Discusses Emerging Fifth-Generation Training Goals

London Fifth-generation pilots will rarely, if ever, fly a training sortie that does not have a synthetic component, Air Combat Command boss Gen. Hawk Carlisle said. “That is the vision of where we’re headed, but we’ve got a ways to go before we get there,” he said, during Defense IQ’s International Fighter conference Nov. 16 in London. “If you look at the surface-to-air systems and the threat density and capability out there, we need to train to that level. It would be cost-prohibitive—and impossible in open air—to create that kind of environment to train against. So could we [train] live, and then embed virtual to increase the fidelity of what these guys face? That’s the intent.”

The Air Force is pursuing the goal of LVC (live virtual constructive) training, where formations of live aircraft fly missions alongside pilots in simulators. “We do live training very well, and we do virtual training very well,” Carlisle said. “What we haven’t done yet is real combined live and virtual. We’re really working towards LVC: how can we pair a four-ship that is in the sims with a four-ship that’s actually airborne, and have them be one flight of eight?” The question is made more acute by the different data-driven demands made of fifth-generation platform pilots. Col.? Trey “Lou” Rawls, USAF liaison to the F-35 Joint Program Office, said during the conference that flying with the JSF’s fusion engine requires a different skillset to legacy platforms. “When I was flying F-15s, I was the fusion engine,” he said. “Each sensor had a dedicated scope, and the job of the pilot was to look at each one and make sense of what that was. What fusion does for you is to take the ‘observe’ and ‘orient’ pieces of the OODA loop, so you can spend more of your time doing the ‘decide’ and ‘act’ portion. The [F-35] fusion engine is a massive mathematical algorithm, and I would argue that, in many ways, in aviation it’s our first hack at artificial intelligence.”