This Guy Has Figured Out a Way to Make Flying Small Planes Safe

• Flying is dangerous, but modern software and hardware can take a basic technology that hasn't changed in 100 years and make it safe.

• A new company called Airhart Aeronautics in Long Beach, California, uses proprietary software and fly-by-wire hardware to limit the mistakes you might make flying.

• Their first safer plane will be out in January 2026 for $500,000, but subsequent planes could cost as little as $100,000. That’s the aim, anyway.

Flying is inherently dangerous. There are a hundred things you have to do exactly right every single time you fly or something terrible happens and everyone dies.

That may sound fatalistic but it’s true. The best pilots are the ones who follow procedures, who can take everything into account all the time, not be distracted, and focus on the immediate task at hand. They also know to check the fuel and oil levels before takeoff, and not fly into thunderheads. And a million other things.

The vast majority of crashes of small planes involve pilots making relatively simple pilot errors. Because it’s such a relatively complex process—compared to say, driving a car—most potential pilots drop out of flight school before they get their licenses.

But there is a way to make flying safer for the rest of us. A new company called Airhart Aeronautics in Long Beach, California, says it has the answer.

Nikita Ermoshkin, a former SpaceX engineer and now co-founder and CEO/CTO of Airhart, has developed a whole new approach to entry-level flying that uses software and hardware to control some, but not all, aspects of flight. Simply put, the system won’t let you do anything dumb that’ll get you killed.

“The big reason flying is dangerous is because there’s only one pilot in the plane and that pilot is a huge single point of failure,” said Ermoshkin. “There’s a thousand things you have to be doing when flying an airplane, what they call ANC: aviate, navigate, communicate.”

Airhart Aeronautics’s new system takes care of many of those things for you, freeing up your mind to work on the rest. With fly-by-wire connections to the control surfaces like ailerons, rudders, and elevators, as well as throttle, the software won’t allow you to over- or under-adjust anything. This will potentially eliminate a lot of crashes.

“We are trying to automate as much of the ‘automation’ part of ANC, or the aviating part of ANC, so that your brain is a lot less loaded by physically manipulating the controls to focus on the things that computers aren’t good at doing but humans are, which is decision-making and risk management.”

They’re also working on integrating the N and the C, so that parts of the navigating and communicating are handled by software.

It’s not the same as an auto pilot. You still have to fly. But the act of flying is reduced to two controls: a push- pull-bar on your right that controls how fast you want to go, and a joystick on your left controls where you want to go. The pedals that normally control the rudder are eliminated, replaced by computer control. Everything else, including the limits of physics and aeronautics, is handled by software and hardware that controls the plane.

Right now, Airhart is adapting a South Africa-built kit plane called a Sling TSi, a four-seat, low wing, monoplane, to use its control system. The Sling TSi is powered by a 160-hp Rotax turbocharged four-cylinder with FADEC 916 iS (Full Authority Digital Engine Control, a computer-managed system that controls the performance of an aircraft’s engine).

Sling says the Rotax engine is “based on the proven concept of the Rotax 915 iS engine, and offers increased power, the best power-to-weight ratio in its class, full take-off power of 160 hp up to 15,000 feet and a service ceiling of 23,000 feet.”

That’s the starting point. Airhart then rips out the dash and adds its own avionics and flight control system, all integrated into a cockpit-wide screen of its own design and manufacture.

“We’ve designed the PFD/MFD system (Primary Flight Display/MultiFunction Display), we’ve redesigned them from the ground up to be much prettier, much more accessible, much easier to use so it’s all touchscreen now.”

The left half is what you might call the instrument cluster, with the artificial horizon, air speed, altitude and other readouts. On the right side of the giant flat screen is your navigation, showing where you are, where you’re going, restricted air space, airports, and things like that.

“Everything you need is integrated into this display,” said Ermoshkin.

There was a simulator in the Airhart office and I got to take a flight. I would taxi and take off from Long Beach Airport (KLGB) and fly to nearby Torrance Airport (KTOA).

“The instrument panel, what we’re calling Airhart Assist for now, all you need to do to take off is push your airspeed controller in—forward to go faster back to go slower,” Ermoshkin coached. “You steer with the stick. So right hand all the way forward, that’s speed control. Pull the left stick back and the airplane will take off once it gets to a safe takeoff speed.”

What’s my stall angle here, I asked?

“So that’s the thing about Airhart Assist, you’re no longer just moving the elevator and ailerons around. You’re commanding a climb rate, a descent rate, and a turn rate. So you can pull full stick back, pull the power all the way back and the airplane won’t stall.”

It prevents you from stalling?

“Exactly.”

So I pushed the speed control forward and, after a few seconds, pulled the flying control/joystick backward and voila, I was flying. There was no worrying about flaps, about stalling by pulling a yoke back too suddenly, about pitching sideways and falling into someone’s backyard BBQ, or even about trying to lift off before I had sufficient air speed. The software in the system took care of all that.

I climbed quickly to 1,000 feet and turned west for Torrance. The three screens of the simulator weren’t full of unnecessary details of buildings or cities below, the better to reduce the speed with which the system reacts. It reacts as fast as your brain thinks is real time.

“In most systems there’s a delay and this doesn’t have that. The delay here is less than 10 milliseconds, so, faster than your human perception.”

That means no hurling!

In a few minutes I was on approach to Torrance (or what I will now call KTOA since I’m a pilot now after 10 minutes of simulator time, ha ha ha). I landed—it was easy—even with a crosswind. The system limited the dumb over-corrections I might have been making the same way traction control keeps you from spinning out in your new Ferrari.

It’s more than software. There are hardware improvements, too.

Right now, the flight surfaces on the single Sling TSi concept aircraft are controlled by rods. Airhart says production planes will have a full fly-by-wire system installed in the planes it sells to the public. Those sales will start in January of 2026. There is redundant everything, including avionics, and batteries. And even if the whole thing goes south, there’s a big BRS Aerospace parachute that’ll save you.

The first planes won’t be cheap. They will cost $500,000, which Ermoshkin says is the going rate for a competitive airplane without Airhart Assist.

One thing: Since it’s technically an experimental aircraft, you have to build it yourself. Or help a professional to help you build it yourself. That could mean as little work as a few weekends. Or more if you’re good at this. The exact number of hours you’ll have to put in is sort of a sliding scale based on tasks versus hours. Ask the FAA. There’s a Sling assembly facility right there at KTOA, one of seven in the US, 19 in Europe.

Another thing: You still have to get a pilot’s license, this is not a Light Sport Aircraft, according to the FAA.

The plan is to sell 50 of this first batch of airplanes. After that, the goal is to make complete new airplanes and sell them for $100,000.

“The big thing we are emphasizing with everything we do is vertical integration,” said Ermoshkin. “For instance, we developed the flight displays ourselves. We do all the bonding and lamination in-house.”

An airplane like this consists of only $2,000 worth of aluminum, Ermoshkin says. The rest is labor and electronics. Streamline that through vertical integration, as he used to do at SpaceX, and cheaper airplanes are possible. His dream is to make flying as easy as driving a car, a car with an automatic transmission. So when you go to Mammoth, or Da U.P., or Upstate, you won’t have to drive, you can fly.

Will they pull it off? Can the average idiot be flying around above your house for 100 grand? It sounds like fun. Let’s hope they do it.