Projects

Autonomous Robot Competition

ENPH 253: Autonomous Robot Competition

As a part of our curriculum for Engineering Physics, students partake in an robotics competition fondly known as “Robot Summer”, course code ENPH 253: Introduction to Instrument Design. While known as “Robot Summer”, students are nevertheless expected to take on a full course load, adding up to a total of 18 credits taken during this time.

ENPH 253 is a four-month long project course. The first two months are dedicated to design laboratories, where we learned to solder, program an STM32 “Blue Pill” microprocessor, design and collaborate on CAD using OnShape, interface with digital signals and build Schmidt triggers using comparators, interface with analog signals through building reflective sensors, and finally, build H-bridge circuit to drive DC motors through a PWM signal produced by the STM32.

Machine Learning Competition

ENPH 353: Machine Learning Competition

ENPH 353 is a project course taken by UBC Engineering students centered around a machine learning competition. Students in this course practice collaboration through git on a linux-based development environment, image recognition through OpenCV, Optical Character Recognition (OCR) from scratch using Convolutional Neural Networks, robotics control through Robot Operating System and simulation on Gazebo.

The competition portion of this course consisted of programming an algorithm for a self-driving car to navigate a simulated environment in Gazebo. The car was expected to traverse on and off road, and navigate obstacles such as roundabouts, pedestrian crosswalks, and other cars. The car was tasked at reading signage along the way, which would give clues to a murder mystery. These clues were to be read using OCR off of a camera mounted at the front of the car, and uploaded to a score tracker during the competition. Each team only had one attempt to run their simulation on-stage.