Automated Gel Turret

The Automated Gel Turret is a toy designed for adults 18+ that uses facial tracking to identify and shoot targets with biodegradable gel pellets. It works by identifying targets using a camera with integrated artificial intelligence to detect objects set by the user, and relays the position of that object to a Raspberry Pi Pico W. The Pico W then uses this position data to determine how much the turret must move to aim itself at the center of the target.

In addition, the Automated Gel Turret also features a manual mode where users can use a graphical interface on their computer to control the turret wirelessly. This allows users to set the turret up independently, or control their gel blasting fun. The turret also includes several safety features including a manual override for the gel blaster and integrated password protection on the user interface.

The turret was constructed using wood, aluminum, and 3D-printed PLA parts, and features a modified off-the-shelf gel blaster. It uses a Huskylens AI camera to identify its target and processes that information with micro-python and circuit-python code on a Raspberry Pi Pico W.

The Turret is mechatronically controlled using a stepper motor to control its yaw (side to side) movement and a 20kg servo motor to control movement in its pitch (up and down) direction.

We built several prototypes before this, starting with a turntable that turned a box side-to-side. As we went, we prototyped each new stage before integrating them together into our final prototype.

From this project, I learned how to integrate various circuitry, sensors, and code, as well as how to control these systems wirelessly using a graphical user interface. This project served as a chance to use my python knowledge, as well as my knowledge of circuits from previous classes I had taken.

One factor that helped us in this project was resourcefulness and good resource management. We were able to source many components without purchasing them new, and a generous donor lent us the use of their Huskylens AI camera. This work kept us well under budget and helped us to get to work faster than if we had waited for shipping for each part.

Our prototype functioned as intended, and met all of the benchmarks, metrics, and requirements that we had set for ourselves, and its functionality. Since we were the stakeholders/clientele for this project we were able to set each of the benchmarks that were necessary for this product to be a success. These included but were not limited to accuracy ratings for the turret's movement, and safety standards for our batteries and gel projectiles.

The Automated Turret project team consisted of Owen Chaffin, Schuyler Chew, Tristan Stephens, and Deven Wells.