I am a student at the University of Illinois in Urbana-Champaign who has been tasked with a Senior Design Project that relies heavily on the communication between a Raspberry Pi and GERTboard. The objective of my team's project is to design and implement an herbicide spraying system and mechanical weeding system that can be paired with an autonomous lawn mower (http://www.coolhunting.com/tech/tango-e5-autonomous-mower.php). The lawn mowing robot that was given to us to use as the project platform is compatible with the Raspberry Pi. As our project began to come together, we realized we needed the GERTboard expansion pack to be successful. (Much thanks to Drew Fustini for kindly donating a GERTboard to my Senior Design Team!)
The system is comprised of a tank of herbicide that gets pressurized by an air compressor. The herbicide tank is connected to a solenoid valve that is controlled by a servo. We already had the Raspberry Pi working, but the Gertboard was necessary in providing the system with relays to control the solenoid valve, air compressor and servo.
Basically, our autonomous lawn mower drives forward until the camera paired with the vision processing software detects a weed. The Raspberry Pi and Gertboard communicate with the solenoid valve, which opens and projects a controlled flow of herbicide onto the weed. A constant controlled flow rate is obtainable by utilizing an air compressor connected to a pressurized tank by means of tubing. When the solenoid valve is open, herbicide is sprayed onto the weed through the nozzle after being actuated to the correct position by the servomotor and robotic lawn mower movements.
Attached to this post is a picture of our internal hardware mounting inside of the autonomous lawn mower. This hardware primarily consists of the Raspberry Pi, GERTboard and Intel Atom Board. The Pi is mostly in charge of the lawn mower's movement. The Intel Atom board is used solely to calculate the centroid of the filtered object (weed/plant to kill) through a webcam. It does not affect any logic decisions made by the Raspberry Pi. Through the serial port connection with the Raspberry Pi, the Intel board receives a centroid request from the Pi and returns the centroid values using RS232 serial communication protocol.
Our next steps consist of designing a mechanical weeding system that consists of an auger-style drill bit that can be centered over the centroid of a weed and drill through the taproot of the weed.