Project Completion Date: December 3rd, 2018
Associations: Purdue Engineering Honors, Purdue
Subject Area: Robotics
As part of the engineering honors program at Purdue, students are required to brainstorm, design, and construct a final robot project in teams of four. Each semester has a unique challenge - in Fall 2019, the robot task was to traverse a simulated habitat by tracking a pre-placed guideline and travelling over several randomized type of obstructions. The robot also had to understand several forms of input, including reflected light and magnetic sources. As a team member, my duties included software development and construction.
The MACRO was designed as separate components that were then integrated at the end of the project. The most important of which, and also the base for the entire machine, was the drive system. Using a set of gears, three wheels on each side were tied together to form distinct "sides" of the robot. Two wheels were in contact with the ground at all times and provided excellent traction control when traversing obstacles. A third wheel was placed in between those two wheels and raised off the ground - this wheel was designed to reduce the possibility that the MACRO could be pinned on a particularly difficult obstacle.
Luckily, the main challenge in the fall semester was very similar to a problem I'd solved in middle school, in FIRST Lego League (FLL). Using a single sensor tuned to look for reflected light, I devised an algorithm to determine the difference between black and white surfaces and then target the robot's direction towards the gradient between the two. In this way, the MACRO was constantly adjusting itself, making sure it was able to keep itself on the guideline at all times.
The software design of the MACRO was built on the principle that the machine had to be completely autonomous during the final demonstration. Routines intended to adapt to any possible situation were created and placed into a massive decision set that would select whatever pre-programmed routines were best for the current scenario.