A trainer doesn’t have to be the fastest, biggest or stealthiest aircraft ever, just predictable, reliable and economical. Air forces nevertheless have a long history of getting trainer programs wrong.

The problem is usually the requirement. The U.S. Air Force’s T-X requirement, seeking a replacement for the Northrop Grumman T-38, has been evolving steadily since the first industry teams formed in 2010-11 – in the direction of a bigger and more expensive aircraft.

As it did with its last all-new trainer, the Joint Primary Aircraft Training System (JPATS) that led to the Textron Beechcraft T-6, the Air Force started with the premise that there were lots of trainers in the world and that there was no need for a new airplane. As happened with JPATS, it has moved away from that position.

That’s not necessarily a bad thing. The rule of thumb is that 70% of the life-cycle cost of a military aircraft is in sustainment; research and development is only a fraction of the remaining 30%. And if you believe that there might be a truly zero-R&D solution to the T-X requirement, you have neither read your history nor the request for information (RFI) that was issued in March. 

Most likely, it was the RFI’s sustained g requirement (a threshold of 6.5 g and an objective of 7.5 g, with 80% fuel at 15,000 ft.) that eliminated the BAE Systems Hawk and the Alenia Aermacchi M346. Alenia’s partnership with General Dynamics broke up after the RFI was released. BAE’s partner, Northrop Grumman, had already announced that it was pursuing an all-new aircraft, with a demonstrator being built by Scaled Composites.

(I’m sure that will work better than the last time someone won a U.S. trainer program with the help of a prototype from Scaled. That was the Fairchild Republic T-46A “Thunderpiglet,” which the flight-test team, for good reasons, nicknamed Eliminator.)

The change has left Boeing and Saab happy. Boeing’s announcement in 2009 that it would pursue an all-new T-X raised eyebrows at the time, and its 2013 teaming with Saab left some people further mystified. Now, Boeing has a one-year running start on Northrop Grumman, and as it becomes clear what the Air Force wants, the relevance of Saab’s small-fighter experience becomes apparent.

Lockheed Martin made noises about its own clean-sheet design, but has since insisted that it is staying with the Korean Aerospace Industries T-50. Some talk about the Textron Scorpion being a candidate, but it would mean big changes to the design and a move out of Textron’s comfort zone.

But will even three contenders show up once the request for proposals hits the street? TX is a big opportunity for Boeing, whichever way Long Range Strike Bomber goes, because of its international connections, involvement in the fighter enterprise (there will be many Boeing fighters around in 2035) and huge support business.

Northrop Grumman will need to take a very close look at its chances in a three-way T-X fight, unless it has some secret sauce – but it has not built anything that looks like T-X since the 1980s. If Northrop Grumman wins LRSB, it will have to focus on execution of that major effort, while T-X will be more important still for Boeing.

There’s still some operator pressure on the requirement. T-X has become more fighter-like because Air Combat Command has embraced the idea of “downloading” – moving training time and events from fighters to trainers. The complexity and cost of the F-22 and F-35, and the fact that they have no two-seat versions, is important – F-22 units already use T-38 companion aircraft. The fiscal 2016 budget includes a companion-trainer package for T-X, including radar, datalink and pylons.

Boeing/Saab and Northrop Grumman have already had to decide whether or not to match the supersonic speed of the T-38 and the T-50.  Supersonic speed is desirable for a companion trainer that has to emulate a threat, less so for a pure trainer: there is exactly one supersonic exercise in the Air Force’s T-38 course. The trick is to minimize the life-cycle cost of adding supersonic speed to the already demanding T-X requirement.

The next question is: What do you call an agile supersonic aircraft with provision for radar and external stores? A fighter. Not one that you would fly head-on into a nest of S-400s, but certainly something that could do air policing, event protection or armed overwatch. That could be useful for an air force worried about rising unit costs and a smaller high-end fighter fleet. Conversely, however, it might be less costly to relax the trainer specification while retaining a few specialized lead-in fighter training aircraft.

Fighter-like performance may expand the market for the contractor, but the risk is that it comes with a fighter-like price tag. Making sure that doesn’t happen is the customer’s job.