Hardware: We’re close to the final design

Last week, we received and assembled the parts of our latest prototype and it’s looking great. It’s quite a milestone for us, so we wanted to share it with you via some explanation of what’s been improved since we last spoke of hardware about month ago. Julien, our Mechanical Engineer will walk you through all the changes and decision that have happened lately, let’s dive in!

The prototype we just finally assembled is the third major version of the drone since the campaign ended. From September to December, we worked on many variations of a simple frame to define the best position, power and propeller size/shape for the motors, this was a very raw version of the drone. We then produced a more polished version in terms of finishes for CES in early January but the insides were still very DIYesque and it took up to two days to put together each drone. Since January we’ve worked on creating a version that will be the basis for industrial production and the result arrived last week. We had already touched upon the improvements we made in this update, but I’ll explain everything in detail here for you to have a global understanding

Reducing bulk (and weight too)

It’s important for us to make HEXO+ as carryable as possible. If you can’t take your aerial camera along wherever you want, it makes little sense to have one. So based both on our assumptions and the feedback we got on our v2 prototype, we set out to minimize the size of the body. The motors position is set in stone so there’s no change in span, but we worked on reducing the overall length by 6 cm, without compromising on the position of the camera (unobstructed footage), which is quite an achievement.

Creating a better architecture (and reducing weight)

Working on the drone’s architecture is a way for us to address several key aspects like the overall weight, the assembly process, the cost of manufacturing, the flight performances or the replaceability of parts. During this iteration, we wanted to shave a lot of weight off, which led us to an important decision, that is to have a shell over the frame and arms that are independent from the frame. This has several benefits:

  • The shell is not structural anymore and it’s thickness can be reduced to the minimum with the material we’re using, ABS. We went down from 2 mm to 1,3 mm thickness when other drone manufacturers are around 1,5 mm for structural shells.
  • The problem with structural shells (the shells is what provides the rigidity to the whole aircraft) is that if they break, you’re good to change the whole drone. We decided to use a structural frame inside a lightweight shell and independent arms so you can simply replace them in case one breaks or gets damaged. With our v2 prototype if you’d break an arm you’d have to change the whole frame. Now it’s just a matter of 2 screws and a weld.
  • The arms are now one single part that fulfills all functions of structure, aerodynamic profiling, motor attachment and cable routing, which means we only need one part instead of 3 before.
  • Working from a frame instead of a shell also helps us working in shorter iterations, because the manufacture time of a shell is twice as much compared to that of a frame.

We also worked on making sure the field of view for the GoPro was even wider than before, to ensure footage that’s completely void of anything but you and the landscape.

Improving assembly and anticipating manufacturing (and reducing weight)

A lot of the work during this iteration was to organize the inside of the drone and all of its components in a way that’s much closer to a finished product than a prototype. That means less double sided tape, gaffer tape and DIY tricks, more screws, cable routing and precise fitting. Moving in this direction is fundamental to prepare industrial production and is also a tremendous gain of time for us. Assembling a prototype now take only half a day compared to 2 days for the previous version (and a projected 40 minutes in the production line). Less time mounting = more time flying and testing. Addressing minute details of the assembly is also a way to reduce weight, because we have had to design every part with two goals: easing the assembly and minimizing weight.

With this version of the drone, we also wanted to be very realistic, so we added all electronics that will fit on the actual product, like and on-off switch, the positioning leds, the usb connector, the gimbal connector and the optional RC transmitter connector. The battery is extremely easy to plug in and replace, sliding into a dock at the back without fiddling with wires or plugs.

Better landing gear
You’ve already seen that in a previous update, but it’s worth mentioning here in details: we went from simplistic, DIY systems for the landing gear to proper feet that can land on any surface, that are very stable and that match the aesthetics of the drone. They’re easy to clip on and off too which means you can pack smaller, replace them easily and why not, have specific feet for specific uses…

As you can see our concerns are mainly around flight time, usability and optimizing manufacturing, which is essentially what we need to create a product that rules and deliver it on time 🙂 Now back to work! Next update on hardware in a couple of weeks.

Julien.

Comments/Discussions

Antoine
Antoine is a co-founder and our flag-bearing CEO, making sure we know where we're going, we deliver on time and we have the most epic team in the biz. Plus tons of stuff that have to do with management, finances, administration, which most people consider wouldn't consider fun. He does.