EDIT APRIL 9: We have had some success on things other than code and motor control today. I have successfully figured out a way to control motor output using IR and also figured out how to use 2n2222 transistors to add motor voltage and increase speed with little weight addition. WE will still be using the arduino micro as it weighs less and our current weight goal is sub 20 grams. We believe this is the minimum we can do on weight in order to lift ourselves and actually fly. This will not be easy. We still need help with code! Anyone you has ever used the MPU6050 for any type of experiment please help us! Feel free to PM me with any design suggestions and input.
EDIT APRIL 8: We had ot ditch the lilypad design. Now we are going with an ardunio mini. After more research the lilypad only outputs 40miliamps to the motors. We need at least 90 to run full capacity. We also tried it and the motors couldn't lift the lilypad alone. So ya...
back to the drawing board.
This is a project a friend and I have been working on per the last few weeks. We are building a proof-of-concpet quadrocopter that runs off of an Arduino LilyPad Simple. Because of its small size it would be very easy to reproduce and make a small army of these things that can communicate and work together like ants. We have considered two different ways of the robots communicating with each other.
1. Robot to robot communication would be done using IR leds that send certain codes out from one robot to the next thus amassing an army in a very short time. Any bot receiving certain commands from one robot or the human commander would be relayed to other nearby quads. The robots would find each other by having 3 IR led receivers on 4 sides of the robot. By tracking which IR receiver it hit first the robot would be able to judge a general "direction" from where the first robot sent the command.
2. The robots use a 2.4 ghz signal and special code names for each robot to communicate with each other. They would use signal strength to find each other. Possibly also each robot has a RGB led with a special clolor and they search for each other using a color sensor.
All of this is conceptual right now. We are still working on how exactly all of this would be accomplished in an orderly fashion. Any advice, help, ideas, or input of any kind will be greatly appreciated. The next step will include revision 1 plans and ideas. including an STL for cad applications. For revision 2 we plan on upgrading to a PCB based design that would weigh less and be more functional for easy building and weight loss. Of course revision two will be much more expensive as we will need to order expensive and super thin PCBs. We will also need a reflow soldering oven for soldering the 32u4s and the gyro chips to the PCBs. We can make revision 1 on our own but if we make it to R2 we may go to kickstarter for funding.
This is not a promise that we will complete this project. My friend and I both lead busy lives and we are not always going to be working on this particular project. Building robots that behave like insects such as ants, bees, or termites is not an easy task. Termites build massive structures blind and without an idea of what exactly each of the other termites are doing around them. Ants also build massive structures and huge complex underground bases by working with each other. Humans do it to. Building robots that an do this for us is a compacted task that is not easy. But please, any input or help or ideas are very welcome! We need your help to complete this! Your ideas matter to us. We are not experts in this field. Noobs rather. The next step will include blueprints and design ideas for the R1 robots. as soon as we get three robots flying and communicating with each other to complete a task we can start moving on to R2. We will also have blog up soon if you want to track our progress. Warning though. The progress will most likely be slow. We can't promise everything will be perfect the first time.
Step 1: Progress N Stuffs (more Coming Each Day!)
Step 2: Revision 1's Basic Build.
1. Arduino lilypad simple.
2. I2c MPU 6050 accelerometer gyro combo.
3. Simple 48000 RPM coreless dc motors.
4. Small 3d printed disc. (If I still had the stl I would give it to you)
6. 40 MAh lithium polymer battery. You can get these on spark fun. They plug into the lilypad.
7. Double sided tape. This stuff hold the arms onto the lilypad and the battery on. Very useful.
You can basically see where this is going if you have ever done electronics before. We use 2n222s for motor control hooked up to analog pins on the lilypad. The MPU is hooked up to 3digital pins on the lilypad. That leaves us with two digital pins for IO. Our first flight will be autonomous. We set it to fly and balance for one second and then land. After that we can get on to communication and control using IR.
Update April 5: Ok guys bad news.... we had the MPU 6050 wired up to the i2c buses incorrectly so we had to resolder everything. but in the process we were finally able to test the MPU. IT works! and is giving us values through the serial monitor! back to soldering my furry friends!
Update april 7th: So I finally figured out how to use PWM on the digital pins and have finally figured out how to control the motors. so now all that is left is to redesign and resolder everything back to its original state. I have also begun research on R2's basic parts such as the 328 SMD chip and the MPU 6050 chip. we have also finalized a way to control the quad using IR. The hardest part is yet to come though. We have to code it. IF any of you guys know how to code this thing please PM me. The basic setup is an I2c gyro that inputs data to the lilypad and the lilypad sends the motors signals on the analogWrite PWM pins with the range of 0 to 255.
Step 3: Code and Updates N Stuff.
//this is a basic little set of commands to tell the
//lilypad to turn on the motors for 2 seconds to see
//if we can get it to take off.
int p1 = 11;
int p2 = 10;
int p3 = 5;
int p4 = 6;
void setup ()
Step 4: Sorry for Not Posting in Awhile Guys Life Happened...
We have made some progress though! We believe we have code that should work well for our purposes and it should get us to a stable hovering state. After that we an add IR control and leds n stuff.