Since school will be over in less than 2 weeks and I will be graduating, then heading off to college I figured that I better get started on my summer project. I have decided that I want to make a fully autonomous vehicle. I have been heavily debating whether I want to make it an aerial vehicle such as a multi-copter or a land vehicle. After much deliberation I have decided on a land vehicle because I already have a really built up chassis and a multicopter would be too expensive for now.
My goals for this project:
Auto navigate to a set GPS coordinates
Do something at a set GPS coordinates
Auto return to home
Switch between auto and manual control with the 3rd channel button on the transmitter
Object avoidance
Automatically perform a maximum speed test
Have a "drag race" mode that will keep the car tracking in a straight line and keep a distance from the car next to it
Log GPS position and speed, g forces, angles
"Launch control" to help from spinning out or flipping over when accelerating hard
What here seems unfeasible?
So here is what I have:
Traxxas Slash 4x4 with 3ch 2.4GHz Rx+Tx, 4S 14.8V 5000mah 40C LiPo, Mamba Max Pro ESC, Tacon 4074 2150Kv BL Motor (Has been clocked at 60MPH+ :D)
Arduino UNO R3
Adafruit Proto Screwshield
Adafruit Ultimate GPS Logger Shield
Adafruit RGB LCD Shield Kit w/ 16x2 Character Display
Maxbotix Ultrasonic Rangefinder - LV-EZ1
Pololu LSM303DLM 3D Compass and Accelerometer
Pololu L3G4200D 3-Axis Gyro Carrier
I have written the code for you. Some of it anyway. I am using Ultimate GPS and Uno. It works for my boat, swamp buggy, wing, quadcopter. Let me clean it up a bit. There were many obstacles to overcome in the coding using interrupts. Conflicts among libraries. I wouldn't want anyone to go thru this again...
fully autonomous vehicle...... (Has been clocked at 60MPH
Probably the main thing you need is a slower truck. The first thing people learn,
from a practical perspective, is R/C trucks and cards are usually way too fast to
make a good robot. Especially autonomous. Although that's always a nice dream.
Is there anything else that I will need?
Yeah, pick just one thing off the list, and do that first. Then, pick another and
add it in. Keep going, one step at a time. By time you're 1/4 of the way through
the list, you'll know where you're at.
BTW, just for reference, somewheres around 1966, Marvin Minsky assigned an
undergrad student a summer project to "solve" computer vision. They still barely
know what they're doing 50 years later.
sbright33:
I have written the code for you. Some of it anyway. I am using Ultimate GPS and Uno. It works for my boat, swamp buggy, wing, quadcopter. Let me clean it up a bit. There were many obstacles to overcome in the coding using interrupts. Conflicts among libraries. I wouldn't want anyone to go thru this again...
Please don't take this the wrong way, but I can't tell if you are being sarcastic or genuine. :~
If you are infact being genuine then I thank you very much!
oric_dan:
fully autonomous vehicle...... (Has been clocked at 60MPH
Probably the main thing you need is a slower truck. The first thing people learn,
from a practical perspective, is R/C trucks and cards are usually way too fast to
make a good robot. Especially autonomous. Although that's always a nice dream.
Is there anything else that I will need?
Yeah, pick just one thing off the list, and do that first. Then, pick another and
add it in. Keep going, one step at a time. By time you're 1/4 of the way through
the list, you'll know where you're at.
BTW, just for reference, somewheres around 1966, Marvin Minsky assigned an
undergrad student a summer project to "solve" computer vision. They still barely
know what they're doing 50 years later.
Yeah, I couldn't tell if he was trying to be sarcastic either, ;-). Some guys are subtle
that way, and some of us have it stamped on our foreheads. "I've got 2500 pages of
source code here that does exactly what you need, and you're welcome to it".
The more of this stuff you do, the more you begin to understand the amount of
work it takes. Robotics is fun, but involves many areas of competence: mechanical
and electrical design, motors, sensors, fabrication, software coding, algorithms,
hardware-software interfacing, plus some semblance of [weak] AI. Then, if it's also
autonomous, you need deal to with software efficiency and real-time operation.
But you learn a lot along the way.
I wasn't being sarcastic. Of course it doesn't do exactly what you need. It's a good starting point though. It navigates the vehicle controlling the speed when there are external factors like wind or waves. Steering to a waypoint. You can turn autopilot on and off with your R/C transmitter.
sbright33:
I wasn't being sarcastic. Of course it doesn't do exactly what you need. It's a good starting point though. It navigates the vehicle controlling the speed when there are external factors like wind or waves. Steering to a waypoint. You can turn autopilot on and off with your R/C transmitter.
Sorry if I sounded rude before. I have just started with playing with some example code from the GPS module and I would gladly accept any code to help me along my way. I wouldn't want to copy it anyways (where's the fun in that?) , just to learn from it.
It's still very much a work in progress. I stripped out the experimental part so as not to confuse you.
The goal is to keep the speed constant, and steer in a straight line thru waves and varying wind.
The difficulty is keeping the boat up on plane and preventing it from spinning out, a difficult task for a human driver.