Our Origin Story

Most great companies have a story to tell about how they started, the trials they faced, the obstacles they had to overcome and how they ultimately persevered through it all to become what they are today.

Flyt is no exception. Having persevered through many iterations of design and engineering challenges, Flyt flew. Not well at first, of course, but we flew. With that seemingly simple undertaking behind us, we went on designing, innovating and building. Pretty much all that it took to bring us to where we are today is chronicled here.


Table of contents

Our 'long history' is, well, long.  What else did you expect?  To make it easier, you can skip directly to a section:

It all started with a dream / The Founding of Flyt / The Mark I

The Mark II / Thrust testing rig / The Mark III / The Mark IV

The Mark 4.5 / The Flyt 16 / The Future

It all started with a dream

No, really, it literally started with a dream.  I had a dream in 9th grade that I could fly to school each morning so I could sleep in longer.  Yup, our origin story starts with wanting to sleep in.

Of course, it didn't look anything like it does now when it was first dreamed up.  Back then it had three propellers each mounted at an angle so you could pull yourself in any direction by increasing power.  I didn't know anything about torque back then and it certainly wouldn't have worked.

Over the next many years, I'd toy with the idea in my head - trying to design it in 3D and rotate it around, figuring out the physics of how it would fly, and making improvements to the design (I stopped keeping track around #41).

Eventually I went to college at USC and decided to major in business and put flying machine thoughts on hold.  After my first year I missed the sciences so I started taking classes to triple major in business, mechanical engineering, and material science.  After loving the physics classes I started to re-engineer the flying machine in my head to work with the wonderful physics I learned about, especially torque (it's what makes something spin).

After a year I finally looked at the class schedule it would require to triple major in these three areas, only to discover they shared very few classes in common and it would take 5 years of 5-6 classes each semester to complete.  So I went back to being just a business major and taught myself some aerospace engineering in my free time, because I still thought my crazy flying machine idea would work.  When I crunched the basic numbers they showed you could generate enough lift to make it fly and I knew I'd eventually have to try it for myself.

After I graduated in 2010 I somehow managed to get a job in management consulting with the intent that after 1-2 years I'd quit to start my own company.  Well, as it often happens to me, reality got in the way and I loved my job so much I stayed for three years while occasionally refining the designs and searching for parts that met the specs I knew I'd need.  Remember, this was before recreational drones were popular so I had to figure out a lot of that stuff on my own.

I recall one day in a fit of engineering I had papers strewn all over my kitchen table (yes, it was from IKEA) and madly drew by hand the plans for my first prototype from every angle and where all the parts would go.  That eventually was the basis for my first prototype.

The founding of Flyt™

After I left my job moved back to the Bay Area with dreams of building a flying machine.  It's funny how people always assume that I grew up dreaming of being able to fly or wanting to solve the world's transportation problem, but the truth is that I just like really hard problems and building a person-size flying machine is really, really difficult (as I later learned).

For six months before I left my job I had a pair of German experimental aircraft engines and 4 carbon fiber propellers waiting for me when I was ready to take the plunge.  So I moved back to Silicon Valley for the simple reason that my father had wood working equipment in his garage I could re-purpose into metal working equipment, plus my parents let me use their garage without charging me rent (a huge plus).  I'm sure you're wondering if this means I moved back in with my parents - no, I moved into a house a few miles away with several friends of mine from college and then went to 'work' every day for 10-12 hours.

When I started I told myself that it would take me about one month to build the first prototype, two months to build the second prototype, and so within three months of starting I'd have succeeded in building a personal flying machine by myself, where teams of talented engineers had tried and failed before me.  This is called either ignorance, ego, or optimism - I'm not sure which.  It's amazing how easy it is to fall into the startup dream and think everything will go smoothly.  This is the period it's fair to call me 'naive'.

The goal at the start was to build a prototype by myself because I didn't want to waste anyone else's time on something that wouldn't work.  I believe strongly in treating others like I want to be treated and I'm OK wasting my own time on an insane project, but not other people's time.  So when I started I told myself that I'd get the flying machine built and prove it worked before finding others to join me.  Spoilers: this is a bad idea and probably makes me a bad entrepreneur.

You may have noticed that I left out telling you when this happened - this was May of 2013 I moved and started full-time work in early June.  Those of you good at math may notice that June 2013 + 3 months does not equal August 2016.  As I said, this was what we call the 'naive' period.

The Mark I

The aircraft engine used on the Mark I

The first 'prototype' was really just a testing rig to see if the basic design would work.  Remember that fit of inspiration on my IKEA kitchen table where I designed everything in one moment of brilliance?  Well, turns out it didn't actually work as well in real life.  Some of the spacing was off and I had to cut out a bunch of things.  It was a mess.

The prototype was essentially a steel frame I welded together, with an aircraft engine mounted to it, which went to a timing belt that turned a set of gears, which then went to another timing belt that transferred power to the propellers.  It turns out, there's a good reason that belts are not used in this manner.

I hung the shredded belt on my bedroom wall as a trophy

There were troubles with pretty much everything, building ducts around the propellers, the drivetrain, even getting the engine started (I had never worked with gas engines like this before and spent two weeks just trying to figure out what everything did).  Eventually I got everything running as expected and started testing, where I learned that it simply produced too many vibrations to work higher than ~50% power and it would destroy a $50 belt every time I throttled it up.  There was even the time I used the wrong type of bearing for the gears (which I bought off of Amazon), and they started to destroy each other in a flurry of shooting sparks.  It was very exciting.

In the end, I learned a lot by building this, namely to use drive shafts to transfer power and be more careful about planning before trying to build something.  Rather than a fully-functional flying machine by the end of three months I had a very loud aircraft engine that would shred belts when turned on.  Physics: 1, Me: 0

The Mark I fully assembled with ducts. And yes, that's a wooden box in front of the engine that was my way of controlling the throttle from a safe distance away using bike break cable.

The Mark II

The Mark II without propellers

Following the belt-shredding adventures of the Mark I, design work began on the Mark II (it's a very clever naming scheme).  I found and ordered some parts from an aerospace gear company, and taught myself some basic CAD to design the drive shafts, and being a former consultant, I naturally turned to Excel to design the basic frame and component locations during the planning process.

The actual construction took longer than expected (shocking, I know) and it was about 6 months until the kinks were worked out and it was fully assembled (complete with more sparks flying out of gearboxes, bent metal bars under the torque load, sheared bolts, and all sorts of adventures.  After I got it running (about 3 months), I slowly and carefully would run tests on each side, bringing up the speed of the propeller in roughly 500RPM increments and running it there before bringing it back down and checking the temperature of all the components.  I wanted to be sure it could handle the forces I was putting on it at higher loads and was always having to adjust the pitch of the propeller lower to get it to go faster and take advantage of the higher engine power.

Around this time I also reached out to the FAA to register my prototype and have a contact to be able to ask questions and understand the rules.  I set up and appointment, and a local FAA representative came out to the garage to talk to me and take some pictures.  Well, actually it was him and two more people who were in training.  We chatted, they took pictures, and he said that he had no idea how to classify this so he'd have to send the pictures to an engineer and would get back to me.  I never heard from him or the FAA again.

My favorite story from this time was wanting to try different shapes of ducts for the propellers and deciding I could 3D print them!  I bought two printers and for over 2 weeks was printing 24/7, including waking up in the middle of the night to start a new print before going back to sleep.  I had to print them in something like 32 segments per duct and then glue them together with epoxy very carefully to keep them aligned and as circular as possible (this is hard for a 44" diameter propeller).  Anyway, I finally got it built and tested it, and things were going well so I decided to go up to full speed where the propeller tip is traveling around 85% the speed of sound.  This works for a couple seconds when the duct breaks apart and chunks of 3D printed plastic going flying everywhere.  Turns out a glue joint had failed and half the duct was sucked into the propeller and absolutely shredded into tens of thousands of pieces.  There were (and still are) a couple dents in the ceiling from flying pieces and it took hours to clean up, but no one was injured and the only damage was the propeller blades needed replaced on that side.  Again, very exciting!  You can see the video on the failures page.

Eventually we got it working to the point that with me sitting in it we could get each side to lift off the ground at full power, meaning it was producing enough thrust to actually lift a person!  I finally hooked up a remote-control system to some linear actuators so I could control the throttles remotely, along with starting and stopping the engines, so it could be safely tested.  Now, this was still in a garage at this time so I had to spent about 2 hours clearing things out and using masking tape to make sure things didn't get sucked off the shelves (they still did).  And don't worry - I also drilled bolts into the concrete floor to secure the prototype so it could only fly a couple inches off the ground.  After much anticipation (I believe it was July 4th, 2014) I finally started up engine #1 and everything was looking great!  I started up engine #2 and everything started to go wrong.  Both engines started to shake and rumble and run incredibly rough when I smelled gasoline (which meant something was really, really wrong) so I stopped both engines and realized my eyes were burning from the gas fumes.  Que a sad and disappointing song.

This problem stayed for another month or two - I tried checking the fuel pressure, spark, everything I could think of.  In the end it turned out that having both engines connected to the same battery was causing a weird feedback loop in the voltage rectifier (I know, technobabble).  The engine though it was turning twice as fast as it was so the fuel injectors were dumping way too much fuel into the engine.  I had to re-wire everything and isolate all the grounds from the metal frame, which is easier said than done.  Eventually I got everything working and was ready to test!

The first tests we got the whole contraption flying and off the ground, but it was definitely not stable and definitely not carrying a person.  The first time it ended rather abruptly when a bath towel came loose and was sucked into a propeller - sending tiny bits of cotton flying all over and making it look a bit like snow.  So of course I quickly shut it off and declared it a successful flight.  The second test we did was better and I made sure to secure all the towels this time, so instead it sucked in some plastic buckets that were sitting high up on a shelf and pulverized them.  It was about this time I decided that testing this monster contraption in a garage might not be a good idea.

After spending about a year of my life on this (and a decent chunk of my savings) I'd learned a few things:

  1. I did not really understand propeller physics

  2. This was not going to work in its current form

  3. Building a personal flying machine is really hard

The Thrust Testing Rig (TTR)

After spending an afternoon to celebrate the fact that the Mark II actually got itself off the ground and drinking a bottle of champagne with everyone helping me, I turned my attention to figuring out what was going wrong.  I decided to build a testing rig that would let me take detailed data measurements of how much thrust the propeller was producing at various speeds and pitches and with different ducts.  So I bought an electric motor and battery pack designed for an electric car and installed it on a rig I built where it could move freely on wheels.

This was great until I realized that the motor I bought wasn't actually producing as much power as the manufacturer claimed and I wasted a good month going back and forth with them trying to figure out why it wasn't producing more power.  It turns out that electric motors are rated for specifications you can never actually achieve in the real world and I was left with an expensive hunk of metal that didn't quite do what I needed.  I worked with it the best I could and found I could extrapolate the data I needed.

Me running a test on the rig (the propeller can't be seen because it's spinning so fast)

I spent many months testing all sorts of things on this rig, using two video cameras to record both the motor data (like voltage, amperage, torque, RPM), and another one that captured the thrust as measured by a scale.  I would then have to go home and carefully sync the videos and go frame by frame to capture data points by hand.  This helped me understand a lot and I was ready to build prototype #3!  Surprisingly, aside from the motor not working right nothing really went wrong here and only two plastic bags were accidentally shredded during testing.

The Mark III

After better understanding the power, torque, speed, and thrust characteristics of the propellers I was using I was ready to build the next prototype which would allow me to actually fly.  I started to design a new, lighter weight frame and improved drivetrain to cut substantial weight and improve efficiency in CAD (so I knew exactly how heavy everything would be).  It would also use variable pitch propellers to change thrust rather than changing the RPM for better control.  This would give the power benefits of gas engines with the stability benefits of quadcopter control systems.  After spending several weeks identifying all the parts I needed, a friend asked me why I didn't go electric instead.  I wrestled with this questions for another week before deciding to scrap a month of work and design an all-electric vehicle instead.

I got the motors and controllers from Slovenia and the only two parts that I kept from the Mark II were the propellers and pilot chair.  There were many technical hurdles, such as the realization after I'd received the motor controllers that the company had very little documentation and half their staff was on vacation.  This led to a full month of frustration trying to get the motors/controllers working correctly without a lot of support given the time zone difference.

The design and construction was slow going - I was sure that this would work and wanted it to look professionally done so I could demonstrate it to people and have them say "wow - I can picture myself flying in that!"  This was not a good idea in hindsight.  Why you might ask?  Simple, it didn't work.  It produced plenty of thrust, but the propellers were so large and heavy that there was substantial lag in the system and it couldn't compensate fast enough to fly. 

If I were to go back and work on it some more now I might be able to get it working (maybe someday), but I also didn't know much about control systems at this point and really didn't understand how they actually worked.  My only experience came from playing around with a toy quadcopter I built out of parts that frequently crashed.  Very encouraging, I know.

The Mark III prototype in final stages of assembly (not all the motors, propellers, and batteries are installed)

Don't get me wrong - the prototype was beautiful!  The pilot was surrounded by carefully designed and constructed polycarbonate safety shields, the power system had carefully bundled wires everywhere, it was made out of special grade of steel, the ducts were carbon-fiber, and the whole thing was painted white so it looked like someone had actually designed and built the thing.  In fact, it looked so good, people assumed I knew exactly what I was doing simply because it looked like what you'd think it might look like.  No one ever questioned me on my skills or if it would work because it looked like it should.  This taught me an important lesson - just because something looks fancy doesn't mean the builder has a clue what they're doing.  Be wary of things that look fancy but haven't been shown to work.

During testing my favorite moment happened on the second test when, as usual, I had to secure everything in the garage from flying around.  I did a better job this time (or so I thought), but it managed to suck in a bag of packing peanuts that was stored somewhere above and shredded the whole thing.  It looked like a snow storm with all the white flying around and took three people two hours to clean up.

Then, there was also the tipping problem.  It would start tipping in one direction due to the narrow left-right base and couldn't correct itself fast (because of the lag in the system).  After a month trying to solve the problem I finally gave up and decided that the Mark III would not fly as I'd hoped.

To be honest with you, this was very depressing.  Imagine spending ten months of your life painstakingly building this with your own hands, secure in the knowledge that it will work.  Then, when the moment finally comes, it doesn't.  So you try again, and it still doesn't.  At the end of this project, I'd spent almost 2.5 years of my life on these projects with few results to show for it - and still no working flying machine.  This is the part about invention and entrepreneurship that frequently gets overlooked: the heartache, pain, and immense frustration when nothing works like you expect.  It's hard.  And I finally understood what all those teams who had come before me to work on the problem of personal flight went through.  I can't express enough how much it sucks when you put in so much work, only to have it fail in front of your eyes.

The Mark IV

In late October, 2015 I was at a decision point - do I keep going after three failures or cut my losses and do something else with my life?  This was a heavy decision after investing so much of my life in this project (~2.5 years at this point).  There were several problems I needed to fix if I were to keep going:

  1. Too much lag in the propellers

  2. Control system not great (and I need to understand it better)

  3. Too much time spent making everything look good before making sure it worked

  4. Hard to troubleshoot problems

  5. Equipment not designed to work with each other

  6. Long time between design, construction, and testing

With these problems in mind I decided I would try it one more time, but with the following rules:

  1. I will spend no more than 2 months from design to testing

  2. Only use things designed to work with each other

  3. Don't include a place for a pilot to sit - but make attachment points that a pilot pod could be attached later (this keeps the weight down)

  4. Keep it as simple as possible to make it easier to build and troubleshoot - nothing fancy

With that in mind I started and worked furiously, running all my calculations in Excel to figure out if the thing would fly or not and if so, for how long.  The first two weeks were finding and ordering parts, so that by Thanksgiving I had the frame mostly build and the parts either arrived or in transit.  There was a lot to learn going from 4 large electric motors to using 8 smaller motors designed for large-scale airplanes.  Everything was different.  So once again, I was building a new prototype that was unlike the previous one and having to learn everything as I go.  Hurray!

The initial CAD rendering for what the Mark IV would look like

This season went by quickly as I was working 12-14 hour days trying to get it ready before Christmas and surprisingly, I did.  The week before Christmas everything was installed and tested and it was ready for its first flight!

Now, given the triumphant nature of that statement you might think it went well.  You would be very wrong.  I had been testing the flight controller on a small quadcopter I built and did not fully understand what the tuning values did.  Essentially I had accidentally made it very sensitive so that it used the full power of the motors to try and correct itself.  Now, before I go further I want you to understand that as much as I love self-deprecation, I'm not an idiot and spend a lot of time thinking about safety in every one of these situations - and that's why I've never had a major accident where someone got hurt.  I secured weights on ropes to the vehicle so it couldn't go flying off and crash somewhere, and only put enough batteries on it to be able to product less thrust than the combined weight of the drone and weights.

So, ready for test #1!  I turn it on and everything is working like expected.  In my head it would nicely lift off the ground and hover gently in the air.  Reality had other plans.  I give it some throttle but when it finally got off the ground I pulled back too much on the throttle so it fell back onto the driveway.  This happens a few times and it sort of tips over on takeoff and almost shreds a propeller a few times.  So I try once more time and it jump off the ground, and proceeds to start picking up the weights and is at a 30 degree angle before landing hard once again.  All in all not what I was hoping for.

So, you'd expect that I fixed the problem and the second flight would work.  But no, you're smart.  You knew from before that nothing ever goes as expected and of course something went wrong again.  The second 'flight' was utterly terrifying!  Picture this with me: I decided to give it a bit more throttle at the start to get it off the ground, and so the 200lb drone jumps about two feet in the air, falls back down, and bounces off the hard surface - but it doesn't bounce evenly.  So, when the computer detects that one side is higher than the other (because it bounced first), it ramps up the power on the lower side to almost full.  This in turn makes that side shoot into the air in a fraction of a second which makes the opposite side suddenly lower, and the computer reacts by making the new 'low side' do the same.  The result is a 200lb machine spinning around at a 45 degree angle trying to level itself before I manage to shut it off (it's at this point I realize I have no way to turn it off other than telling it to go to zero throttle).  It finally slams back into the ground, throwing the lithium batteries violently off the frame (which potentially can explode or catch fire on sudden impacts), and shredding the tips off a couple propellers on the driveway in the process.  Again, utterly terrifying.  I hope this is obvious, but don't ever try this at home!

Here's the video for your amusement:

If things are not failing, you’re not innovating hard enough
— Elon Musk

By this time I was starting to live by the Elon Musk quote, telling myself over and over that I must be innovating really hard because things are always failing!  So we made several improvements to the prototype, for instance we moved it onto grass so that it couldn't bounce, secured the batteries in a much better way (packing tape), and I got to learn all about flight controllers to understand what was actually happening.  We knew it could work, but it was tricky to get it to actually work.  It would jump off the ground but it certainly wasn't stable and we couldn't get it to fly, for instance, 1 foot off the ground and stay there.

Finally, test #8 I think we got it flying in the air for 17 seconds and stable!  It proved that it was possible to build a large drone that flies, but the real question now was how much it could lift.  I built a payload bay that could be attached underneath the vehicle and add weights to it.  On test #28 I got it to fly with 100lbs of payload with no problems.  A few days later we ran several tests with it carrying 200lbs of weights, but it didn't fly very well being so close to its limits in how much weight it could lift.  This is very early February 2016.

The Mark IV on a grassy lawn for testing

The Mark 4.5

Now that the Mark IV would fly and could carry weight, but not quite enough for safety, I decided that the obvious solution would be to simply upgrade it.  So I ordered everything we need to upgrade it from 8 -> 16 propellers to roughly double the thrust (for you engineers reading it, the efficiency lost from stacking propellers is roughly canceled by the weight savings from not requiring the existing components to double, for instance, by using the same number of batteries). 

Here's the Mark 4.5 on a lawn where we'd run some tests. You can see one of the safety weights sitting in the lower-right corner that would get tired to the frame on some rope.

Now the observant reader may notice that we're only in early February and it's September 2016 when this is being announced, so what happened?  Our idea was great in theory, but in practice there was a shortage of the motors we needed and after waiting for 2 months with no word, we had to find an alternate supplier.  Then, they had problems producing the number of motors we needed and the whole upgrade process which should have take about 1 month ended up taking from early February through early-May (3 months).

I won't bore you with the details, but increasing the thrust had several unexpected side effects that led to problems in the flight computer.  It took another month of testing to figure out that our existing flight controllers wouldn't work and that made a whole mess of problems whenever we tried to carry much weight.

There were a lot of fun moments (read: scary, utterly terrifying) during this time, like when it was more powerful than the calculations had predicted and it started to fly for a couple seconds while carrying the safety weights tied to it and almost flew off.  After that we tied two tent stakes to each weight so that it really couldn't fly off.  It also seemed to enjoy flying perfectly normal about a foot off the ground, then suddenly shooting into the air at full power before I could throttle down and let it fall back to the ground.  Again, very exciting times!

Finally in early June we got it working with a new flight controller and some custom adjustments so that it would reliably fly while carrying ~200lbs of weight (that was our goal for a typical person-sized payload).  I wrote in my notebook the day before we got it working "It's OK to fail - I'll learn something new!"  That's the sort of attitude you need to have on this type of project, because it's easy to let the constant failure get you down.  Now, in hindsight this was a huge deal because it could conceivably lift closer to 400-450lbs and still fly, but at the time it was just another development after a long string of failures. 

We ran a lot of tests once we got it working (45 flight tests in total), eventually getting it to the point we could make it hover about a foot off the ground and within a 1-2ft radius of its starting point.  This steadily improved to the point we could fly it for about as long as we wanted, I think our longest was a minute and a half, until a motor failed and I smelled electrical burning.  Understandably we landed it immediately, even though it was apparently flying just fine with one motor not working.  Long story, but it was a smaller motor we were temporarily using while waiting for a back order of the normal sized ones.

We kept pushing and trying new things until one day we turned up the PID values (it's a control systems thing) too high as I was demonstrating it for a friend.  It flew great for ~1 minute before going into an unstable oscillation and crashed into the lawn at a 55 degree angle.  Luckily we had a remote kill switch installed on it by now so the moment it started to oscillate by more than ~10 degrees I just cut all power to the motors and let it fall.  That way the worst that could happen was some vehicle damage.

It was here that I decided it was no longer reasonably safe to test in that area and I really needed a better place.  Also, I should stop messing with the tuning values and stick to what worked.  That lesson only cost about $200 of damage to the vehicle.

There's one more thing I want to explain here.  It's easy to fall into the trap that everything sounds bigger and fancier than it was.  But here's the truth, this prototype was built in a garage, and this 'testing lawn' was in the backyard of the house.  We would have to take all the batteries off, tilt the prototype straight up, and carry it by hand around the house and through a narrow gate, before re-assembling everything and hooking it back up.  It was usually an hour-long process just to move it and set it up for a test.  Then the batteries take forever to charge so we could only do a limited number of tests each day (around 5), plus we'd only test between the hours of 10AM-3PM so we wouldn't bother the neighbors with the noise.  Additionally, if it was a windy day that added risk to the test and sometimes we decided not to test at all to be safe.  So this was the other reason we needed a better space - we'd simply reached the end of where this setup could realistically take us.  Lastly, testing it with a pilot pod would put the propellers above the fence, and I didn't think the neighbors would appreciate it.

Notice the weights (and bricks) stacked below. That's the payload bay where we usually added 200lbs of weight.

The Flyt 16™

Ah, so here we are in the story to the part you actually care about - the thing where people fly!  After our last crash (yup, we're not afraid to talk about our failures because that's how we learn), we decided to try and find a place to rent where we could test this in a safe manner.  After a while of searching and contacting various places we discovered that short-term industrial rentals are not a thing, because, who else in the world needs industrial space for a short time?  So this was quite the setback.  It ended up taking ~5 weeks before we were able to rent an aircraft hangar for a month from a very nice gentleman who was willing to let the crazy inventor rent a hangar to get video of the giant flying multirotor.  There were a couple other people I found who were also willing to help, but the timing didn't quite work out.  My biggest surprise was that not a single person told me "no, you're crazy."  Honestly I expected more people to be scared or otherwise pessimistic but everyone was super nice and positive.

So, in early August 2016 we moved into our new home for the next month!  Above all, our focus was on safety here.  We had an elaborate testing protocol figured out where we would incrementally test things:

  1. Get it flying by itself, no weight, with tethers to make sure the baseline vehicle is working

  2. Fly it with the payload bay, 200lbs weight, still tethered

  3. Test it with the pilot pod, no weight, but with tethers

  4. Test it with pilot pod, incremental weights up to 200lbs, with tethers

  5. After flying it many times with no problems in step #4, test several times without tethers

  6. Once it's consistently working like it should, finally put a person in it and fly, but always low enough to the ground that if something goes wrong they wouldn't be hurt

This setup was great in theory and we were already on step 3 by the second day when we ran into an unexpected problem - it would almost tip over when we tried to fly with the pilot pod attached.  Essentially, it would start to tip over when we would throttle up because it wasn't ramping power up to the propellers evenly.  This led to some clever engineering where we simply used duct tape to extend the legs and make it less prone to tipping - problem solved!  Except not.  It still tried to tip over.  Sigh.  So we then did what any normal person would do - we bought a yoga mat, cut it up, and wrapped the feet in it so there would be a large squishy layer to help it level itself.  Shockingly - this didn't fix the problem either.  Then we tried just taping mini exercise balls to the feet to provide cushioning, but the end of the metal was sharp and punctured one of the balls before we could even try.

So for the next few days I constructed three separate solutions to fix the problem: 1) I bought a replacement mini exercise ball and steel bowls so it would have a really squishy leg that could act as a stabilizer, 2) longer leg extensions (thinking the first ones just weren't long enough), and 3) shock absorber feet to really, really give it squishy legs.  Now I should mention that the original design called for shock absorber feet (because old me was apparently smart) and somewhere along the way this had gotten dropped.  So we try the first one with the exercise balls and it sort of works.  The vehicle scoots along the ground but I'm too afraid to give it enough throttle to fly.  On to option 2!  This one again sort of works, but it's still trying to tip and I have no idea why.  So we try the third option, and in the process of mounting them we end up putting too much pressure on it and break two of the legs.  These were aluminum welded instead of steel which is a lot harder to do, and the aluminum bends much easier.  So two days of welding work is down the drain and we're almost out of battery power to test that day.  Things are going great.

Now, this is August 10th, and we only have the hangar for a month.  Which is why the timing was particularly bad that I was on vacation from the 11th-16th in Utah to visit national parks (this was planned months earlier).  Luckily, during this time I was able to order some more supplies that could be shipped while I was gone.  When I returned it took one day to weld new shock absorber legs out of heavier steel that would not bend.

On August 18th we installed the new legs and what do you know?  It flies!  Now this is very exciting - especially because when it takes off it starts to fly right at me.  A several-hundred pound prototype flying towards you is terrifying in any situation, even more so when you're not expecting this to happen.  We end up adding weight to it and keep testing, and it keeps wanting to fly towards where I'm sitting for these tests, but I get better at controlling it each time.  We keep testing it for several more days trying to figure out the problem before deciding we just need to do a quick fix and move on.  So we jack up one side, re-level it, and that mostly fixes the problem.  This is the startup mentality.  We also had a motor controller short out in a flurry of sparks during a test that required some repairs and upgrades to make sure that didn't happen again.  Then we also had our battery voltage display stop working so we couldn't tell how much power was left in the batteries and had to manually check after each flight.  Plus, this limited the flights to ~30-45 seconds to make sure we didn't drain the batteries too low.  Again, the startup life is very exciting!

Now I'll let you in on a little secret about batteries: they suck.  Not literally of course, but figuratively.  They're heavy, they don't carry much energy, and that means we can only run an average of 4-5 tests per day (each one lasts ~1-2 minutes).  Then, once we're done testing, the batteries take a couple hours to re-charge before the next test (currently around 8-9 hours of charging for each day of testing).  This is largely because we don't have a secondary battery set and have limited charging equipment.  This was made even worse when we were running a test at high power levels, and noticed when I went to check the battery voltages that 1/3rd the batteries were hot and 'puffed', which means a very bad chemical reaction is happening inside the battery and extremely flammable hydrogen gas is being produced.  The short story is that a third of the batteries could potentially explode or catch fire.  Wonderful.  So we quickly took them all off (not really, it's a very slow process) and set them as far apart as we could in case they caught fire.  Oh the excitement!  I should note, that we were operating the batteries well within their specifications, because the rest of the batteries with identical specs were fine and had no problems.  Mostly likely this batch had a manufacturing problem.  The end result was that we lost ~15-20% of our battery capacity, which further reduced our ability to test each day.  Things are going great!

One of our brave test pilots

Finally on August 26th we feel comfortable enough to do the first human flight!  And so on hangar test #45 a person sits in it and gets off the ground!  It wasn't a great test, I was terrified to take it more than a couple inches off the ground in case something went wrong, but a person flew nevertheless.  Over the next five days we had several more people fly in it for numerous tests, including one where a motor shorted out, but that was entirely my mistake because I didn't notice it had gotten stuck and finally got hot enough to melt the insulation.  Again, I'm not afraid to admit that mistakes happen and each time we learn a lot from it.  In fact, I think that it shows just how well this was built that all these problems happen and yet nothing bad ever occurred during testing. So finally after a little over 3 years of work I've built a flying machine that can carry a person and fly with them (relatively) safely!  Yes, you may applaud now.  Oh?  That would be strange to applaud in front of your computer?  Ok, just take a moment to be impressed then, because it wasn't easy!  Thanks.

You might be wondering at this point why we didn't just find an open field and test there.  This is a good question.  There are four primary reasons we wanted to find an indoor testing location:

  1. No weather to worry about

  2. Lots of equipment to move that doesn't fit in one load

  3. Safe environment where if something goes wrong no one else gets hurt

  4. Legality - if we're outside we fall under FAA rules, whereas inside we're not in public airspace

I cannot emphasize the importance of the last point enough.  The last thing we want to do is make the FAA mad at us considering we need to work with them to make this a reality from a regulatory standpoint.

The future

Our original plan was to build what was essentially a flying pod after the Flyt 16 that would improve flight time and safety, but as is frequently the case, plans change.  Instead of jumping straight to pods that fly us around, we're pivoting to focus on the entertainment market in the short-term, similar to how Tesla focused on luxury cars at first even though their goal all along was mass-market electric cars.

We're hard at work fine tuning our latest 12 rotor aerial flyer called the Liberator. With specially designed light-weight motor mounts and carbon fiber panels, the Liberator looks to be the hero of the next chapter of our story!

Thanks for reading, I hope you enjoyed our story as much as I enjoy sharing it!

Forget what is, create what will be
— Ansel Misfeldt

One more thing...

So there's a not-so-secret but still not advertised section down here for those who read all the way to the bottom (or let's be honest, just scrolled all the way down).  In case you didn't read the page called 'A Guy in a Garage', Flyt™ currently only has one employee (me, my name is Ansel - it's nice to meet you!).  This whole project is my passion and I've had to do it mostly by myself.  Now, I've had tons of help from friends and family over the years but I'm the only one working on this all day, every day since 2013.  That's been really hard, and at various points I have made efforts to find others to join me, but for one reason or another it doesn't work out (commitment, skill, availability, etc.).  Ironically, I'm a really social person so I go a bit stir-crazy spending most of the day alone physically building these prototypes.  I owe a huge debt to all those who have helped me on this project in one way or another and could never have done it without their support.

I use terms like 'we' and 'our' because I know that I did not do this all by myself.  I firmly believe in the power of working with other people and never want to be someone who needs credit for their work.  The credit isn't important - the end result is.  I hope it's even more impressive to think about how it was mostly done by one guy literally building all this in a garage by himself.  Still, I'm way past the point of what just one person can do and looking to take it to the next level.  If you think this project sounds just crazy enough to be interesting and think you can help in any way we'd really appreciate it.  Perhaps you'd like to join us as an employee, or can help us find funding, or connect us to potential customers - we appreciate anything.  Check out the help us page.

And lastly, I'm a guy who majored in business, quit his job to work alone in a garage all day to solve a challenge that's vexed brilliant minds for decades, and actually succeeded in building something that can fly a person.  If you think I'm a crazy idiot for doing all this - you're probably right.