The car will never become smarter than the pilot - TASS Opinions

Evgeny Fedosov — about when to wait for sixth-generation combat aircraft, whether neural networks can replace pilots today and thanks to what Russian MiGs could fly with pierced fuselages

The creation of every Russian fighter, bomber or attack aircraft of any generation began with the development of the concept by the State Scientific Research Institute of Aviation Systems (GosNIIAS). On the day of the 78th anniversary of the Institute, we recall the past of combat aircraft construction and its prospects.

Compatibility issues

The aviation industry as a technical branch arose at the beginning of the twentieth century. Both civil and military directions were science-intensive from the very beginning: it was necessary to solve both problems of aerodynamics and problems of strength, to create aircraft engines and control equipment. In this regard, several major scientific institutes of the aviation industry were established in our country in the 1920s and 1940s.

However, they focused on the construction of transport aircraft, not military ones. But such machines have a lot of features, primarily because of the aviation armament, which in the 1930s consisted of cannons, machine guns and conventional bombs. For example, during the discharge of the latter in flight, vortices are formed when the bomb bay is opened, so the ammunition immediately falls under the high-speed flow, which can turn it over or press it against the fuselage. Such phenomena must be prevented. To solve such problems, the Institute of Aviation Armament NII-2 (now GosNIIAS) was established in 1946. He helps the construction of military and civil aviation already at the stage of ideology, the development of the concept of the aircraft.

Among the first problems that the institute's staff had to solve was the compatibility of weapons and aircraft. The first generation of jet aircraft, which included the MiG-15 and MiG-19 cannon fighters, was subsonic, although the MiG-19 could switch to supersonic in some modes. When the MiG-21 was created, it was already conceived as supersonic — there were problems of supersonic aerodynamics and the operation of weapons at supersonic.

Then navigation and sighting systems appeared. And the appearance of the first guided missiles led to the development of a weapons control system. For example, the first air—to—air missile was guided by a radar beam - the locator on the aircraft created a conical beam scan, and the receiver on the rocket "searched" for the energy center of the scanning cone, which stood in the tail of the rocket.

The first locators appeared in aviation in the early 1950s, missiles with homing equipment appeared later. A whole class of weapons arose, which was assigned to NII-2, whose specialists analyzed and solved all emerging scientific problems and helped the creators of weapons.

Without air conditioning, but tenacious

NII-2 had a whole scientific direction for working on the survivability of the aircraft. The results of this work could be observed during the Korean War in the early 1950s, in which Soviet MiG-15s fought on the side of North Korea, and US pilots fought for South Korea on American F-86 Sabres. The latter characteristics corresponded to our MiG-15 and MiG-17, even resembled them structurally. But for 10 damaged Sabres, on average, there was one moment. The fact is that Soviet engineers created the aircraft so that it would continue flying with a punctured wing and holes in the fuselage and return to base — for the Sabre, one shell hit was enough.

This became possible because we understood how the fragmentation field of the enemy's air—to-air missile warhead is formed. We developed recommendations for designers: how to book a cabin, protect fuel tanks, electronic circuits. 

Another telling fact: during the Iran-Iraq war in the 1980s, Iraqi pilots preferred to wait for departure in the cabins of comfortable French Mirage aircraft with working air conditioners, and to carry out their flights on MiG-21. It turned out that the MiG, with the same flight properties, is many times more survivable than the Mirage. By the way, the Tu-22K was also sold to Iraq. One day, an American-made Iranian anti-aircraft missile "Hawk" flew into his open bomb bay and exploded inside. However, the explosion did not affect the power frames and the plane landed. After that, a rumor went around that the Russians were doing "something fantastic."

Mathematical models

Civil issues entered the sphere of interests of the Institute in the 1990s. Safety, not efficiency, is at the forefront here, because 200-300 people are sitting behind the pilot. To create well-thought-out aviation systems, you need to take into account how the pilot interacts with them. For this purpose, back in the 1950s, the first semi-natural modeling stand was created at the institute. Roughly speaking, it was a mock-up of a cockpit with a three-stage gimbal, simulating the movements of the aircraft in roll, pitch and course. Today, these tasks are solved with the help of the "Electronic Bird" modeling complex — this is a kind of "digital airplane". He mathematically repeats the aircraft to the smallest detail. There is no aircraft "in hardware" yet, but the "mathematics" that describes all its movements, functions, and actions of its equipment is already there.

And then we began to remove mathematical models and replace them with real hardware. Moreover, the most difficult task was to put a real radar with a real fairing that would take into account all the effects of refraction when a radio signal passes through various media. 

Now the Institute is building new modeling complexes for civil aviation for MS-21-310 and SJ-100 aircraft with domestic equipment, and is carrying out import substitution of equipment.

Network and Intelligence

Modern aviation systems — and not only aviation systems — involve automating processes and integrating them into network-centric systems that control all operations. They are responsible for the processes of intelligence, communications, management of combat groups, tactics, strategy, and aviation functions. This is called integration — and this is another scientific direction of GosNIIAS.

And although network—centric control is still in the process of formation, we began to implement its principles already in the third generation of combat aircraft - MiG-23, Su-24, Tu-22. Aerodynamics with variable geometry appeared, and the number of control functions began to grow.

Now it is no longer possible to lay down a combat aircraft just because the designer scratched his head and thought: "Yeah, there will be such and such wings, such and such a fuselage." First you need to understand the concept of an airplane as a system. It forms the basis of the terms of reference. To understand this concept, it is necessary to engage in combat operations in a certain sense. This is how cooperation takes place with military specialists who develop tactical techniques.

Manned aviation will remain forever, because no one has surpassed the human brain yet. And we have an intuition. Our brain has great resources that manifest themselves in a critical situation — and the more complex the combat operation, the more important "intelligence" is. Therefore, we are trying not to replace a living pilot with a car, but to create all the conditions for performing complex combat missions.

I am convinced that it is impossible to exclude a person from a combat operation. A car will never be smarter than a pilot. Yes, it is much faster: there are experimental machines that perform billions of operations per second. But where intuition is needed, you need to make a quick decision in a desperate situation, the car will not be able to do anything. 

There is a lot of noise around artificial intelligence right now. The more complex a neural network is, the more opportunities it has to process information. Yes, airplanes are becoming more and more "intelligent", but this is the usual automation of the control process.

"It takes a man to figure out tactics."

There is an opinion that the air group should be mixed and consist of drones and manned aircraft. Such a mixed park, in principle, can exist. Because the drone is the same size and speed as manned aircraft, it can work in a group, perform the work of a slave.

The appearance of air—to-air missiles with a flight range of 300 kilometers or more also makes the use of drones relevant. They can attack targets in the near zone, but in order to understand the tactical situation after a strike, a person is still needed. Moreover, he must work in the far zone so as not to run the risk of being shot down.

But here the problem of artificial intelligence arises, because this drone still has to make independent decisions within the boundaries of its responsibility. This is where some kind of intelligence is really needed, there must be some other solution - not just a neural network. Which one? — a question of questions! After all, a drone can be trained in certain typical techniques in the same way as any pilot. However, the enemy will come up with some kind of cunning maneuver - then the device will become useless.

Sixth generation aircraft

Currently, we are thinking about the concept of a sixth-generation aircraft, conducting exploratory research, and exchanging opinions with military specialists. Such an aircraft should appear sometime by 2050, but already now we need to understand what armed conflicts of the future will become. I think they will really be hybrid, because there is a powerful information struggle, political and economic, as well as a direct military confrontation. 

The Institute believes that the sixth generation aviation group will be mixed (manned aircraft and drone), while the drones will operate in the near zone and will be able to work as suicide bombers (kamikaze mode). 

All previous generations of aircraft were created with the participation of GosNIIAS and military institutes. The concept of the fifth-generation Su-57 aircraft was also created by us, and we are still working on it.

As for the development of aircraft design, we know only one thing: going further along the logic of complication is a vicious practice. Airplanes are becoming more complex from generation to generation due to the growth of functions. Their weight is growing. If the MiG-17 was no heavier than 7-9 tons, then the MiG-21 weighed 10 tons, and the MiG-23 weighed 20 tons. Then there are the MiG—27 weighing 20-30 tons and the Su-57, which overcame the bar of 30 tons. And the bigger and heavier the plane, the more expensive it is.

All paths lead to a multifunctional aircraft — there will be no mass aviation anymore. Mainly due to the fact that airplanes are very expensive, and the economy is the decisive factor.