After years back in school this was it, this was the class that would differentiate me from the Computer Engineering students; CMPE 118: Introduction to Mechatronics. Crafted by Professor Gabriel Elkaim from a course at Stanford, this is arguable U.C. Santa Cruz's toughest undergraduate class (at least within the engineering department, and where else would the most difficult class be?).
Five frantic weeks are spent getting ready to build a fully autonomous robot; learning an event/service framework required for the final robot, designing, testing and eventually fabricating an analog infra-red signal filter, the use of various motor controllers such as H-bridges, and learning how to use Solidworks in conjecture with a laser cutter. The next five week are spent designing and building a robot that must complete several tasks autonomously. While the minimum specification to pass the class was a solo run in the lab, the highlight of the class (and one of the highlights of my academic career) was the head to head competition, in which our robot came third out of fourteen.
Jaiden King
Jeffery Lo
Jordan Cox
Teams were semi-randomly assigned. Self reported strengths in the areas of electronics, mechanics and software were matched to create hypothetically equal teams, or at least, no teams with an obvious weakness. Yet, in every class there seems to be one team where all the members name's start with the same letter.
By far the most challenging and rewarding class I took at U.C. Santa Cruz; the huge work load required a lot of time to be spent in the lab, the four different lab partners before the final project meant everyone worked with five or six other students during the course, which developed a real sense of camaraderie between all the students, culminating in the final competition which felt more like a celebration, not just for the completion of this class, but, with many students in the class graduating a week later, it was also a grand sending off experience.
Designing, prototyping and building the infra-red signal filter was a good example for this class. The filter was designed, then prototyped on a breadboard, and finally required to correctly identify a specific frequency signal from a sequence of signals at a minimum range. This was a difficult and particularly time consuming process, but nothing compared with the next lab which required the design to be transferred to a single, relatively small, perf-board. My partner and I ended up spending over 20 hours over a weekend soldering our filter. We had observed other students soldering many components, or even the entire circuit, before testing and debugging, and that this had caused many problems. So, we decided to test every op-amp based component when it was completed, fixing any problems and ensuring it worked as expected, before moving onto the next.
Our robot was required to move to the firing zone after being randomly placed (orientation and position) in the starting zone, while being able to navigate around randomly placed obstacles without seriously disturbing them. Once within the firing zone our robot needed to hit a target placed in the firing zone of a field mirrored to the left of this diagram (with accompanying randomly placed obstacles).
This video shows our robot meeting the min-spec requirements.