Embedded Files
Facial Recognition
Facial Recognition
Our facial recognition model was an engineering project focused on creating an AI model capable of accurately identifying the person shown in an image. Each student developed their model individually using Google Colaboratory. The image dataset we used consisted of photos we took and labeled ourselves. There is a more detailed documentation for this project attached to the left.
video of skittle sorter.mp4Skittle Sorter
Skittle Sorter
Final Project Overview
The final version of our Skittle Sorter includes several key components working together to sort Skittles by color automatically. Here’s a breakdown of the parts and their functions:
Funnel:
The sorting process begins with the funnel, a 3D-printed part designed using Onshape CAD software. It has a wide opening at the top and narrows down to a tube, guiding the Skittles into the sorting mechanism.Grabber:
The grabber is made of circular cardboard pieces connected together and attached to a servo motor. It safely transports individual Skittles toward the sensor by rotating.Servo (Grabber Servo):
This servo rotates the grabber to move the Skittle toward the sorting hole. It is controlled by a microcontroller called the Raspberry Pi Pico, which runs the sorting code.Stopper:
A unique addition to our sorter, the stopper is a servo-driven piece of cardboard that temporarily holds the Skittle in place. This pause allows the sensor to accurately detect the Skittle’s color before releasing it down the ramp.Sensor:
The sensor detects the Skittle’s color by measuring RGB values and sends this information to the Pico. This data enables the code to predict the Skittle’s color for sorting.Ramp:
After color detection, the servo connected to the ramp moves it to align with the correct bin. The ramp guides the Skittle into the designated bin.Bins:
These are color-coded containers where the sorted Skittles are finally collected.
Code Summary
Code Summary
Stopper Sweeper Servo Code:
The grabber servo rotates for a fixed duration to prevent jamming and then stops.
Sensor Code:
The sensor uses preset RGB parameters to identify the Skittle’s color based on the data it receives from the hardware.
Ramp Servo Code:
Based on the color detected by the sensor, this servo moves the ramp to specific angles corresponding to the correct bin location.
Reflection Questions
Reflection Questions
How did this project compare to others you have completed?
This project was more complex than any previous ones due to the many small, interconnected parts that all had to work perfectly together. While the prototype took a very long time to build, the final version was completed much faster thanks to the experience we gained.What is one challenge you overcame, and how?
A major challenge was Skittles getting stuck while moving from the grabber to the stopper and sensor. After trying many small adjustments without success, I decided to remove and completely rebuild the transfer mechanism. Surprisingly, this fix took only about 45 minutes and resolved the problem.What are you most proud of?
I’m most proud of our unique stopper mechanism. Although it caused some complications, it improved the consistency of color sensing. We developed it as a workaround since our top servo didn’t have angular movement, and it added valuable functionality to our sorter.
Process Documentation
Process Documentation
Process documentation is a detailed record of all the steps you followed to complete your project. It explains how you designed, built, tested, and refined your robot, along with the challenges you faced and how you solved them. This can include:
Planning: Initial ideas, sketches, and goals for the project.
Building: Materials and tools used, along with assembly steps.
Coding: How you wrote and tested the code to control the robot.
Testing: What worked well, what didn’t, and adjustments you made.
Reflection: Insights gained and improvements for the future.
Page updated
Report abuse