At one point I tried to do a drone startup. It was a long project, and lasted about three years. It didn’t end in a fireball or anything like that – heck, that would have at least provided some Youtube videos. It ended with a silent passing on: killed as it struggled to open its eyes. This was about two years ago now. I don’t regret the experience at all, but it’s taken me awhile to write about it because I’ve been so busy with other things, but I think it’s important that I get this out there, if not for anything other than my own sake: we all cope with the knowledge of our smallness in different ways. If you’re out there, cooking up some idea in your head with your friends: maybe this story is for you.
So let’s start with the people. My compatriots were Kyle Blemel and Shawn O’Neil. Both immensely talented individuals in their respective fields: manufacturing/design and aerospace engineering. Kyle basically handled the building of the project – I did design, business planning, CAD modeling and project management, while Shawn did the reality checks and number crunching. He put together some amazing spreadsheets that I’d like to share with you folks later in this article – hopefully they’ll be useful for you.
We had the idea to develop a drone that would directly counteract what we saw as a “me-too” approach to drone design in the consumer market. There was (and still kind of is, but less so now) a real rush to make quadcopters for virtually every task – cargo carrying, mail delivery, camera platforms – you name it, there was a quadcopter for it. This to us smacked of a market that was saturated with people who didn’t truly understand aeronautical design: quadcopters were the hot new thing, they were the latest technology – so obviously they surpassed everything else before them, right? Not so much.
Aircraft are designed for missions. Those mission profiles are determined by requirements for speed, endurance, lifting capacity, and maneuverability. Quadcopters are great at maneuverability, and they’re decent at speed: so they make for great camera platforms. But when it comes to endurance, well – they’re awful. They will literally never be as good as a comparable fixed wing aircraft, because all of their lift (the force that keeps aircraft in the air) comes from power. So if you want more endurance, you need to add more batteries, so your power can last longer – you can’t make the wings bigger, because the quad doesn’t have them. But like all aircraft, they fall prey to self reinforcing spiral: more endurance means more batteries. More batteries means more weight, which means more power required to lift, which means less endurance. There is a hard cap to the endurance and load-carrying capabilities of a quadcopter at any given size, and this cap will remain fixed until battery density improves.
So in this physics problem, we saw an opportunity for a high-endurance aircraft that could carry a relatively large load. Instead of going for a fixed wing aircraft, we decided to go for a parafoil-based platform. These flexible fabric wings could theoretically provide for the maximum amount of cargo capacity possible: with relatively little weight allocated for the lifting mechanism (a fixed wing’s wing still weighs something), a parafoil based aircraft could feasibly lift 500 lbs into the air for only maybe 20 lbs of structure. Pretty enticing, right? It helped that Shawn had also worked on a similar project for his thesis. Check out the slide from our pitch deck below for how it stacked up:
However, we immediately ran into problems deciding on the mission, and these problems would dog the project until the very end. It didn’t help that we were also trying to tackle two problems at once, too. You see, the pitch just wasn’t for this cargo-carrying drone: it was also for a larger augmented/virtual reality game based around dogfighting and racing drones. Players would compete in real-world tournaments, flying actual drone aircraft. Potential players at home could join in on the action via a free simulator they could download, bringing the experience of flight to every living room. Our company would make money on in-game microtransactions and aircraft kit sales. If this sounds familiar, that’s because someone else has ended up doing it. But hey, you can check out some of my pitch deck slides below:
So now we had an unclear mission: this drone was supposed to become a camera drone for this larger augmented reality video game. But it was also supposed to be a standalone product in the consumer market. But it should also be able to carry cargo. I think in retrospect we should have either picked one or the other. Since we had no development talent (none of us were programmers), we should have focused exclusively on either a high endurance camera drone or a high endurance cargo drone. But we were beset with the challenge of finding a market: it wasn’t until much, much later in the project, around the final year, that we started to see mainstream consumer use of drones for things other than photography. Too little, too late.
Troubles with Prototypes
Contributing to the lack of direction was the dislocation of our team. I was in New Jersey at the time, then in Boston. Kyle was in upstate New York. Shawn was in the Seattle area. I really believed it was possible for a disparate team to build a physical product, despite not even sharing a workspace. We used cloud-based CAD to review designs. I juggled the production of the prototype in one city, and the design analysis in another city – all the while being in a third city myself. I think this lack of a central workspace was the biggest mistake that doomed us from the start – we didn’t prototype anything for nearly two and a half years. Why?
1) We had hardly any expertise in the electronics side of things at the time, so a lot of our formative years were spent reading instead of experimenting.
2) I had a very broad vision for the company, based on a stair stepped approach – one step would lead to the next, until we built our whole product (a VR/AR drone line and game system). What this meant was that the first step was constantly changing to try and make a “good enough” product that would get us to the next step, rather than focusing on making a profitable initial product only.
3) We could spend hours discussing what we should or could do, but had no way of just trying out ideas, and were hesitant to spend any money (fresh out of college) on ideas because we didn’t know if they would work or where they would go.
There’s probably a few other reasons that an MBA could pick out, but this lack of simply doing things is what really killed the project, in my opinion. Had we started building prototypes early on, we would have learned a lot faster, and been able to take advantage of the nascent market. But due to our slow pace, we never got anywhere. Yet we still worked. During the time this project took, I had switched jobs twice, and eventually landed as a freelance designer. I was really hoping for this to “take off” (pun intended). I put together a pitch deck outlining the whole project, and built funding timelines. I combed through the Small Business Administration website and followed their guides on what we would need to run the business and present to investors. You can download them here. My favorite one is the “Time to Repayment”, which shows how many units we need to sell at what price point to break even. We did our homework, and we built what we thought was a compelling business case.
As all of this was being put together, we started developing the prototype. In developing it, we found a few major issues that challenged our initial assumptions. The first was the difficulty in locating a parafoil of a suitable size. We could find some R/C parafoils that could lift about 1.5-3 lbs, and then we could find some full-size parafoils for carrying people – but there was literally nothing in between.
So of course our first prototype wasn’t going to be full size, and we went with the Hobbyking parafoil. Once we started designing around it though, we encountered another problem: parafoils are a pain in the butt to work with. They have a ton of strings going everywhere, and they’re difficult to roll up and keep untangled. Then, to launch the thing, you have to make sure the parafoil is carefully laid out so that it can fill with air properly on take-off. Again, our lack of initial direction came back to haunt us: if this was supposed to be for the consumer market, people probably won’t want to go through the hassle of setting this thing up to launch. If it’s supposed to be for the commercial market, what was its mission? We found that the military does use basically something similar to what we were designing, with the CQ-10 Snowgoose, and they require it to be launched from the back of a truck with people running beside it – not exactly easy to use.
Ok, so we had a super tight weight budget, and were still struggling with a direction and target market. But we were going to make something, damnit. So we pushed on, and after several late night calls, we were really close to producing a prototype. I drove to upstate New York to help Kyle finalize the assembly, and attempt our first test flights. We had two days: I drove up on Friday night, and had to be back Monday morning for work. This trip really drove home the importance of being able to work in a shared space: it took us almost 8 months to build roughly 60% of the prototype remotely. When I was able to work with Kyle in the same room, we were able to complete the prototype and test fly it in just a day and a half.
The Test Flight
The prototype we ended up building was a strange combination of a proof of concept and design piece. On the one hand, it was supposed to try out our control scheme for controlling the parafoil. On the other hand, it was supposed to show the general concept of a parafoil-based drone that had a large swappable “cargo pod” that could be changed out for different missions. There was some confusion here in my communicating this to my teammates, and I think that was probably in large part because the overall mission still had yet to be defined – so this prototype was also relatively undefined. To Kyle and everyone else gathered, the red cargo pod (which was made of foam and weighed maybe 1/4 lbs) was the cargo – and not just a conceptual stand-in for what was to be the cargo later. So, that probably could have been explained better.
Anyway, we finished the prototype, and went out to test fly it. We were excited – I mean, this had been a day coming for 3 years. So we drove out to the park, set up the drone, laid out the parachute, and gave it full throttle…
“This trip really drove home the importance of being able to work in a shared space: it took us almost 8 months to build roughly 60% of the prototype remotely. When I was able to work with Kyle in the same room, we were able to complete the prototype and test fly it in just a day and a half. “
…and nothing happened. The plastic wheels we had 3D printed wouldn’t roll on the rough asphalt pavement. Weight wise, we were probably a pound and a half over our budget, and couldn’t get the aircraft to accelerate enough to make up for the difference. Our prototype didn’t even have the glory of a crash – it just didn’t move. We tried to run it behind a skateboard, cut out some weight – everything we could think of. But nothing would convince the little thing to budge.
Defeated, we packed up and went to share a beer. We named our little red friend “Scarlett Dronehanson”. I packed up my stuff, and went home. After that, we never really went anywhere. The concept was too flawed, our direction too indistinct, and after three years, I know I was just tired of working on the project. People were already way ahead of us, and the overall concept, the augmented reality game, was a pipe dream without any development talent.
It’s not like we didn’t profit from the experience. It was exciting, and we all learned a lot of great skills – I learned a lot about project management in particular. We also had bought some equipment for the purpose of prototyping that we ended up sharing with each other. So it’s not like we came out of the project with a total loss. But as a startup, it was stillborn.
I don’t regret the project, and I don’t regret doing it. I’m glad I got to try doing a startup, and got to make an attempt at making what I still think is a wonderful idea a reality. If I were to summarize our mistakes as advice, it would go like this:
1) If you’ve got a great idea, define a business case before you try to make a startup out of it. Otherwise, it’s just a hobby.
2) Prototype early, prototype often. Don’t be afraid to spend money to build things.
3) If your project requires some sort of talent that you don’t have, your job as project lead should be to get that talent: no matter what. Don’t run with only one leg. Pay someone if you have to, offer services – just get the talent your team needs.
That’s it for me. I hope this helped someone. I tried to cover as much of our story as I could. If you’d like to view my entire pitch deck, you can find it here.