n_ermosh

The control system knows what the rudder should be at all times for coordinated flight in the air, and keeping the nose and ground track aligned when near the ground for landing, performing the slip necessary when landing in a cross wind.

I'd love to chat through your questions directly--feel free to book a call with us through our website (just mention this HN post) and I'm happy to talk through everything

You can revert to a "direct law" in an emergency should multiple faults occur that reduce the capability of the flight controller. However, there's a question of is that the right move--if a pilot isn't regularly practiced in flying in a direct control, its probably more dangerous to revert to direct law than to say, pull the parachute, so it would be up to the pilot to determine the best course of action. The system would still be fully "by wire" though, there isn't a mechanical link between the pilot and surfaces/throttle. But, in such a mode, the flight controller is bypassed and the yoke directly commands the actuators, reducing the number of fault points between the pilot's input and the actuator.

We've flown early versions in an test aircraft that we have and are currently integrating the latest version of Airhart Assist (seen in the video) into the Airhart Sling prototype to be flying and giving demos by the end of the year.

I think there's definitely a "tried and true" and "if it aint broke don't fix it" mentality. Combine that with the fear of trying something new in the face of the FAA and regulatory barrier, many will say it's better to just take the risk on older stuff than make something new and better

Well, not exactly. In case of an engine failure, the system still has power from the batteries to control the actuator and glide the plane (in fact, it will automatically switch to flying at the best glide speed to extend it's range) to a safe landing, whether it's at an airport or in a field. But if you are over some trees, mountains, or somewhere else where a landing is impossible (which is a problem in today's planes too) the parachute is a final backup. the parachute is also the backup in case of a complete system failure

we should talk!

We are exploring how to make a subscription model work. It's not possible in an experimental category airplane, but our future models where they can be rented (legally due to FAA requirements) will make that possible.

Airhart Sling is just the first step too. yes, it's expensive, but it gets the tech into today's market. Our mission is to scale the market to develop a plane that costs <$100k by bringing in a new wave of people into aviation--those who don't fly airplanes today for the reasons I've discussed.

There's an interesting dynamic here--we (the industry today) are more okay with flying rickety airplanes from the 70s before flying something built with more modern engineering and production techniques.

> As I understand it the idea here is to expand the market and spread certification costs out over more planes.

exactly.

> There are not enough people interested in a personal airplane at any price to support the costs to bring a new one in the air.

our thesis is that this isn't true. we've seen glimpses of this in the past 10 years that haven't been successful, but have shown that there is a wave of people who would get into GA if it were safer and more affordable. our mission to make it so, and the Airhart Sling is just the first step

fair criticism--I should be more specific than just "green".

our first model will still be a gasoline ICE airplane, but we are building a limited number of them to get our tech out there and get a new wave of people excited about GA. out future aircraft will be able to utilize our fly-by-wire/simplified controls to fly more aerodynamically efficient airframes that will use less energy, whatever that source is. We can do this because these airframes would be too unstable for a human to fly by hand, hence the need for fly-by-wire/simplified controls in the loop.

today, we don't have a good alternative to ICE for airplanes that also meets the mission profile of the vehicle. I personally think hybrid electric is at least the next step, with hydrogen (though that has it's own challenges) coming after.

yep--we want to reverse that trend. And you are probably right, a lot of those 70s planes still fly today. hopefully we can replace them with better alternatives

We'll still encourage and require the same proficiency checks that pilots do today.

Our hope is that GA pilots will no longer need to do things like train stall recovery. We are moving into an era of aviation where aircraft (not just airplanes) will be complicated enough that computers have to be in the loop to handle things like that, because it allows us to create more interesting aerodynamic aircraft that are more efficient and have better performance.

regarding something like AF447, the immediate answer is we have the ballistic parachute as a final backup in case the pilot is unable to land the plane, for whatever reason. Realistically, it would depend entirely on the exact situation our plane was in, what systems have failed vs which haven't, and the pilots actual skill level

The core innovation here is the fly-by-wire/simplified control scheme. This doesn't exist in any GA aircraft.

As our system has been developed so far, it has negligible impact on the speed and usable load of a base TSi, costing $500k for everything.

we've definitely thought about that! the force-feedback fly by wire makes it really easy to replicate the feel of flying the real thing very well. What you see outside though will be fixed, so maybe for instrument training it would be good, though I'm not sure about using the actual plane vs recreating the cockpit in an indoor simulator

this is definitely more of the "roadster". we will be building a limited number to get it into peoples hands as quickly as possible, with future models being lower cost, more efficient, and hopefully even better in terms of control systems, avionics, and autonomy

We've talked to many and the complexity of flying a plane is a big factor--because the knowledge that if they mess up stick and rudder means they have a high chance of losing control and crashing scares them from flying. Our system helps alleviate that fear.

it is indeed a problem. however, Sling (who we are working with on our first model) is also developing a "Whisper" version of their airplane to make it dramatically quieter (https://www.slingpilotacademy.com/airport-noise-reduction/#1...). We hope to adopt the tech into the Airhart Sling as well

> 150kt Cruise, 4h endurance - 400kg "useful" load for pax/luggage @ 4h endurance / batteries - 3-axis AP that'll do coupled approaches + autoland - BRS, FIKI, Oxy/Pressurized - $100k fly away

Totally agreed. Our long term roadmap has basically this exact product on it. However, anyone who doesn't know about GA will have no clue what this even means. We want more people to get into GA so that there's enough of a market that building such a plane becomes a sustainable business.

> The Ercoupe tried this exact approach in the 30s in one of the big GA booms and it just wasn't borne out.

The Ercoupe was a great idea for it's time, but we can do so much more with the technology we have today that we didn't have back then

> a production aircraft that's cheaper than $50/hr to fly would.

yes! our long term goal is to get there, but we need more people flying airplanes today to generate enough scale to actually build such an airplane and also stay in business. That's our approach, get more people who can afford to fly doing it, then use that to drive down the cost for everyone.

Today's batteries aren't quite good enough to hit the same performance. But developments in the hybrid electric space look promising. Battery technology is improving though, so in the relatively near future, it very well could be

a combination of altitude, commanded descent rate, and commanded airspeed transition the control system from flight to landing, flare and touch down, and into ground control mode.

Multiple sensors, actively modeling the flight dynamics to detect anomalous sensors/behaviors (i.e. is what we are measuring what we expect to measure), and BRS as a final backup.

yes I should have mentioned actuators as well. the actuators are internally redundant in terms of computation and power and will fail silent and not move if they disagree. Then, each control axis has 2 sets of these servos.

I say axis, rather than control surface, because for something like ailerons, we can have one go neutral in a fault case and the aileron on the other wing still operates with some reduced control authority. So even though each aileron has one servo, the roll axis has 2 and it maintains redundancy

The core flight simulator is x-plane with some of our own stuff on top

we won't want to lower the bar in terms of the pilot's ability to stay safe. But we do want to lower the barrier to entry so that more people can learn to do it, do it safely, and enjoy all it's benefits.

In your case, you do start to approach the point where time-wise it's about equal. But, 30 minutes to your nearest GA airport is on the long side, most people (300M) live within 15 minutes of one (don't have a direct source for that--it's based on analysis we did of locations of airports and population densities)

We want to see prices go down too. That will only happen if more people want to fly airplanes and the market grows. So, we are trying to make it more accessible to people who wouldn't otherwise want to fly.

redundancy in sensors, computers, and actuators reduces the probably of a system failure due to random errors/failures and dissimilar computer systems help reduce the probability of a common mode software bug causing a failure.

I can't speak for what will or won't happen with MOSAIC, but the current proposal bakes in SVO so we (and a lot of other companies) all hope that stays.

Yes, our first product is experimental and has a 51% requirement, but Sling has a great factory build assist program that will help with that. Typically, electrical systems aren't included in a 51% build and Sling builders don't usually touch the avionics (when doing a factory assist), so we hope that stays true.

The things that exist today are good steps in the direction, but we are pushing the boundary further with fly-by-wire, where these is very little innovation.

We are starting in experimental, working with Sling Aircraft for the airframe, factory build assist, etc to prove out the exact product that we want to take to certification in the future.

In the 70s when a lot of this system was designed, the industry was selling 10x the number of GA planes it does today, so I imagine 10x more were flying as well. So we have the capability for an order of magnitude increase. We'll need more controllers, sure.

But, you are totally right, we need to move away from voice based ATC to more digital systems that allow a management of more aircraft, and allow the GA aircraft to manage themselves through a "peer to peer" type ATC system, but like we have with our cars.