MicroFly is a quadcopter packed with WiFi, 6 sensors, and 3 processors for ultimate stability -- and fits in the palm of your hand.
Have you ever seen a quadcopter? Its a wonderful piece of engineering that requires an extreme sense of balance to pull off. All four rotors have to work in unity to keep the machine from tipping over. With the power of modern technology, hobbyists around the world can build their own quadcopter using different components from different manufacturers. Most hobby quadcopters are large and requires lots of power to operate.
I would like to do something different.
Instead of designing a quadcopter that you piece together, why not make a quadcopter from scratch? While at it, why not make it as tiny as possible?
I've been inspired by the crew at bitcraze.se, where they are building a simple quadcopter at a tiny scale; every component is carefully selected to be mounted on the circuit board. The circuit board itself acts as a chassis for the quadcopter, and custom 3D printed components are used to mount the motors. The reason? To make it as small as possible, without redefining the quadcopter. I would like to do the same, but with some important differences.
So what makes my design different than other open source quadcopters?
In my design, I plan on using embedded WiFi, so the quadcopter can be controlled from a computer on a standard wireless network you might have at home. As a result, the quadcopter can go as far as your wifi signal stretches. This quadcopter is small, which inherently makes it excellent for indoor use.
The quadcopter will have redundant IMUs (Inertial Measurement Unit) to ensure a stable flight for the entire battery life. The Microfly will also run on a small lithium-ion battery that can be recharged from the USB port on your computer.
My design is not meant to be simple, its meant to challenge the minds of everyone, including long-term engineers. I want to go all out and see how much I can fit on a 4 inch circuit board. This is no "arduino"; it's meant to teach how the circuit works as opposed to providing a toy without explanation.
Why I need funds to continue
I have been working on the MicroFly in my free time for the last couple weeks, but my project has hit a serious financial bind. Due to the circuit's complexity, I simply cannot afford to build the quadcopter prototypes without the help of others. I need tools (a hot-air gun, oscilloscope, and a bench-top power supply) to test the design; I need funds to purchase the initial set of prototype boards, and possibly a second round of prototype boards.
My current status
I have most of the schematic for the first prototype completed, and now moving on the the component placement stage, which can take some time. Above is a 3D rendering of the circuit board shape that will be used in the first prototype. Remember, this thing is LESS than 4 inches wide. I'm only a few days work left on the initial schematic, and below is a preview of the IMU circuitry.
UPDATE 1: I have placed most of the components on the board! There is some tweaking to do still before I start routing the traces, but this is a good start! Here is the preview of both the front and back of the board.
Future plans for the project
Every component of the MicroFly will be completely open-sourced. This includes the schematic, the PCB layout files, the drill files, the PCB shape, the software, and the bill of materials. I plan on selling the MicroFly in the future as well to hobbyists who don't have the tools to build a MicroFly themselves.
Thank you for all your kind support, and I look forward to seeing this project reach its true potential.
The first prototype run will be using the Atmel AT32UC3B range of microcontrollers for the main "processor" and the Atmel ATMEGA range for each IMU to help offload the sensor processing from the main microcontroller.
The Kickstarter set of production grade board will be hand-built, not in a production line, but the funds raised along with personal funds into the project will go into getting fees covered for the eventual production line. The beta software is defined as complete when the copter flys stabily. Production grade software will continue to be updated, but the Kickstarter boards will be complete when battery management, power management, and better stability code is complete.
$5000 is the minimum needed to get the project going. I will be matching up to $2000 additional out of pocket some of the expenditures. Since this Kickstarter is more for ramping my idea to production grade, everything is going to be hand-built, hence the limit on the total quadcopters that can be purchased. Each board will cost about $225 in parts (including PCB, components, and 3D printing material, stencils). The rest of the costs factor in tools needed for the build such as a hot-air gun.