How can you make a cool Arduino robotics project even cooler - by adding a Raspberry Pi component to it! This is exactly what we did when preparing for our demo booth at Open Hardware Summit in MIT earlier this year. Raspberry Pi is powerful enough to execute a computer vision algorithm such as object tracking and Arduino is very handy when it comes to robotics applications and motor control. Simulink can implement both the computer vision algorithm and the motor control algorithm using library blocks. We decided to put it all together into Rasduino - a robot that tracks a green ball in its field of vision. MakerZone has a detailed blog post about the Arduino control algorithm on the Rasduino.
Here's a video from one of our test runs:
It was our first time exhibiting at the Open Hardware Summit and the demo was well received on the exhibit floor. It was great to see the enthusiasm about robotics projects in students and hobbyists. We got lots of exciting ideas on how to expand the demo to include other cool features such as face detection capabilities. Moreover, number of instructors asked us how they could recreate the demo. All in all it was great fun putting the demo together and sharing what we have with other hobbyists.
Lesson learned: In terms of recreating the demo, the Simulink models can be reused very easily, but the hardware itself is not easily replicated. In the next version of this project, we have decided to use a robotics platform like the AAR robotics platform to make it easy for others to reuse our efforts.