Inspiration
Back in my hometown despite there being dedicated landscapers to take care of plants in the city, I noticed some areas could get neglected. Sprinklers would break and be left without repair, and furthermore, sometimes plants would be overwatered! These are all issues we aimed to solve with STRAW-E.
What it does
STRAW-E regulates the watering of plants. Currently, STRAW-E can deliver a precise amount of water to any plant pot.
How we built it
A robotic arm regulates watering by lifting a water hose up and down, with water being gravity-fed from a raised water tower. We used two servos to construct the robotic arm. To control the servos, we created an Arduino script to run on our ESP32 computer. Additionally, we implemented a Wi-Fi interface that allows any connected program to control the servos. To demonstrate this functionality, we developed a remote Python script. Our goal is to enable cloud-based algorithms to optimize and control the robot.
Challenges we ran into
- Implementing the exact controls for the servos required a lot of fine tuning. Since we were using continuous rotation servos, we had to get the timing just right for the rotation rate to be correct.
- We had some difficulties controlling the water flow, as sometimes our piping would leak!
Accomplishments that we're proud of
- The ability for our arm to run remotely through Wi-Fi. This opens the gate for cloud-based algorithms that can really optimize the robot!
- We came up with a simple, low cost system for holding water that simply relies on gravity through a "water tower" and the control of one servo motor. When the plant should be watered the servo motor moves the arm down to let water flow and when the plant should not be watered we move it up.
What we learned
Through this project, we developed new hardware skills, particularly in programming servo motors and working within the Arduino environment. More importantly, we learned how to transform an idea into a functional physical system composed of multiple integrated components. These key working parts include the straw mechanism in the water pipe, the water tower, and the robot arm. This experience deepened our understanding of robotics and hardware.
What's next for STRAW-E
- To continue our progress with STRAW-E and really transform our cities there are a few key components to integrate. First, we need to create a four-wheel platform that can enables the robot to water any plant in a wide area. This platform should be robust enough to operate on any terrain.
- Next, we plan to implement computer vision and navigation algorithms. These algorithms would allow the robot to recognize plants in its landscape and navigate towards them to provide water.
- Finally, we want to take watering to the next level with reinforcement learning and an array of sensor data. Sensor data, such as temperature, humidity, and soil moisture, will optimize a reinforcement learning algorithm that will always provide the ideal amount of water to any plant.
Log in or sign up for Devpost to join the conversation.