January 26, 2021: At the end of 2020 the U.S. Air Force successfully tested Artu, a new software system for managing sensors on a U-2 reconnaissance aircraft that acts like the “weapons officer” or GIB (Guy In the Back) that became common during the 1960s to handle sensors and weapons system while the pilot concentrated on flying the plane. Over the decades the software available to the GIB, or just the pilot in single seat aircraft like the F-117, became more capable and able to operate without much human intervention at all.

The term Artu is a play on R2D2, the Star Wars “astromech” droid whose main purpose was to serve as a copilot on single-seat fighters. Artu was touted as taking control of the aircraft sensors and making decisions on which sensors to use for which task. This could be mainly seeking out enemy air defense sensors to be attacked by other aircraft or defensively to detect threats to the aircraft. These are life-or-death decisions being entrusted to an AI (artificial intelligence). What most people outside the flying community don’t understand is that Artu is evolutionary, not revolutionary.

As aircraft became more capable since the 1970s it became common to entrust more and more functions to software systems that could be overridden by the pilot but most of the time just did a lot of work the pilot, or pilot and GIB, could not handle manually and still operate as a fully capable combat aircraft. One of the earliest examples of this was the first stealth warplane, the American F-117. Developed during the 1970s and entering service in 1983, the F-117 was a single seat aircraft where the pilot had to handle a lot more tasks than any pilot could deal with. That was because the F-117 was a light bomber that used laser guided bombs as well as sensors to detect targets as well as potential threats. Most important was the flight-control software. An important aspect of the F-117’s stealth was its odd shape, that made it difficult for radar to detect or track. This odd fuselage also made the aircraft aerodynamically unstable. The F-117 not only needed a fly-by-wire control system, rather than the older manual/hydraulic controls, but flight control software as well that did most of intricate manipulation of the flight controls. Without this software the pilot could barely control the aircraft in level flight and had no time to operate the elaborate fire-control software. The use of electrically controlled flight surfaces (flaps, rudders and such) in a fly-by-wire warplane was relatively new in the 1970s and one of the key aspects of keeping the work on the F-117 secret was to hide the fact that a much more advanced fly-by-wire system was being developed. The Russians in particular knew that such a flight control system would be needed to make a stealth aircraft like F-117 possible. Russia had also discovered the effectiveness of an odd shaped fuselage to provide stealth, but thought suitable flight-control software was way in the future. When they became aware of the F-117 they knew the Americans had solved that software problem and managed to keep it secret.

Since the 1980s all aircraft manufacturers have sought more and more capable flight control software. Another goal was more elaborate and capable fire control software. This sort of software is not seen as daring as fly-by-wire flight control software because failure of the fire control system does not make it difficult, or impossible to fly the aircraft. Yet fire control software has also benefited from the AI-based software developments. Israel has often taken the lead in this and one of their more recent developments is the Fire Weaver fire control system that takes data from existing sensors on tanks and other armored vehicles as well as artillery and warplanes, and rapidly (within five seconds) lets vehicles, warplanes and artillery know which available target each combat system should fire at. This eliminates a common battlefield situation where too many weapons fire on some targets while other targets are not initially fired on at all. Currently, tank crews and artillery spotters (troops who call back to tell artillery which targets to hit) have manual procedures for picking which targets they should fire at. That often works quite well, especially during a situation where a tank unit encountering the enemy has an opportunity to fire first. Fire Weaver automates these decisions and makes more effective choices more quickly. The troops and pilots can override the Fire Weaver selected target but tests have shown that Fire Weaver is usually quite effective in selecting the best targets for each tank, artillery unit or aircraft.

Fire Weaver was easy to implement in the IDF because the Israelis have already been providing their troops with better sensors and battlefield networks. For example, in mid-2019 three Israeli firms, responding to an IDF proposal, showed off their versions of the proposed Carmel Concept for future armored vehicles. Three different armored vehicles; the Merkava 4 tank, Namer IFV (Infantry Fighting Vehicle) and the Eitan 8x8 APC (Armored Personnel Carrier) had proposed versions of Carmel installed. Carmel involves several existing technologies plus proposed new ones that would turn an armored vehicle into a “combat system” that could operate with, a crew of two or a robotic vehicle operated remotely (like a UAV) or autonomously to benefit from more information about where friendly and suspected enemy forces were. This information would often be delivered in real-time. This sort of thing provides a tremendous advantage in combat.

The same sort of thing has been happening with war planes and here the U.S. has been in the lead. The best example of that is the F-35 fire control (and flight control) software. Pilots new to the F-35 realize this immediately because one thing the F-35 does extremely well is use automated flight controls that allow the pilot to carry out maneuvers that would require a lot more experience in older (F-15. Su-30) aircraft but are much easier for an F-35 pilot. The more experienced pilots know a lot more useful maneuvers than new pilots but because of the adaptive F-35 flight control software it is much easier for new pilots to master an unfamiliar maneuver. The best way to explain this is the experience of British carrier pilots who formerly flew Harrier vertical takeoff and landing aircraft and were now using the vertical takeoff and landing F-35B. The British pilots were told that difficult carrier landings that could be terrifying in a Harrier, which U.S. Marine Corps pilots also used on small carriers, were surprisingly easy with an F-35B. As British pilots began carrying out landings on the new British carrier they were pleasantly surprised. The F-35B flight controls automatically adapted to all the rapidly changing wind and carrier movement variables and allowed you to land without a lot of stress. Handling the F-35B in general was much easier, and safer, than the Harrier. Hovering, for example, required a lot of continuous effort and attention from a Harrier pilot. In the F-35B the pilot could fly the aircraft to a position and hover and the aircraft would remain where it was flown to without additional effort by the pilots no matter how much the weather changed.

All this ease of flying enables F-35 pilots to concentrate on something that does still require a lot of decision making by the pilot; stealth management and threat management. The stealth characteristics of the F-35 make it more difficult for radar to detect it. How the pilots fly in a combat zone can improve the effectiveness of stealth. That is done by learning to manage the flood of “threat management” data that F-35 pilots have access to. By being able to concentrate on stealth and threat management F-35 pilots achieve what has been the key element in air combat since 1914; getting in the first shot. From 1914 into the 1940s the key to success in air-to-air combat was knowing how to fly into a position where you would see the enemy first and carry out a surprise attack. The earliest of these tricks was the World War I tactic of trying to have the sun behind you as that made it more difficult for the enemy to see you coming. Another tactic was trying to get higher and out of sight for as long as possible until you could dive on the enemy aircraft in a high speed and unexpected attack. In effect, “stealth” and the resulting surprise was always the key to victory. The F-35 was designed with that in mind. The F-35 radar stealth and maneuverability isn’t as good as the F-22, but the F-35 “situational awareness” is much better. Pilots who have flown the F-22 and F-35 always note that and point out that, in the hands of an experienced pilot, it makes the F-35 a more effective aircraft than the older and more expensive F-22.

In that way the F-35 was designed to have “affordable stealth” and much more effective sensors and electronics. The F-35 stealth is much less expensive than that in the F-22 and initial Israeli combat experience over Lebanon and Syria indicates that the stealth and internal electronic countermeasures more than make up to for that. The passive sensors and “sensor fusion” software of the F-35 also appear to be working as advertised. In the cockpit, the pilot has one large (20-inch diagonal) LCD showing all needed aircraft data with more showing on the pilots JHMDS helmet visor. That is all very well, but as with the very capable F-22 it wasn’t performance that limited procurement but excessive cost.

What the F-35 flight management software and situational awareness demonstrate is that the usual measures of a superior fighter aircraft (speed, maneuverability) no longer matter as much. An F-35 is more likely to see the other aircraft first, fire first and be more aware of the changing battle situation than enemy pilots in, on paper, faster and more maneuverable aircraft.

Even when the F-35 is hit and damaged the flight control software senses the damage and automatically flies differently to compensate for the damage. That takes a lot of stress off the pilot who can concentrate on threat and stealth management to complete the mission and get the aircraft back to base. Another important aspect of the F-35 is that its flight control and threat management software is built to be constantly updated by pilot experience. As more pilots fly the F-35 and experiment with different techniques the software is updated to become more capable. Those updates require more attention to post-change testing. That’s because there are so many interconnections within the flight control software. Those have to be tested to prevent unexpected results when the pilot is most vulnerable to that sort of thing.

Artu for the U-2 is another application of the F-35 flight control/sensor control software. The media depicted Artu as an electronic co-pilot but that is nothing new. Software controlled flight assistance systems began appearing in the 1970s. This was mainly in the form of an audible warning system, often in the form of a human voice warning the pilot of a looming flight emergency that required immediate pilot action. Soon some nations were expanding the tasks these systems could handle and that evolved into Artu on the U-2 and similar systems on all combat aircraft. The F-35 is currently the most advanced example of that.