• Start software early. It’s easy to get bogged down in hardware, especially since you may have several iterations there. If you write software in parallel, you’ll be ready to debug as soon as your hardware is built. Otherwise, you’ll likely be very crunched for time on the software. Remember, in mechatronics, no matter how good your hardware is, it won't do anything without software to run it, and it won't do anything well unless that software is solid.
• Just because you built a particular circuit for a lab this quarter, don’t put it off until the end. We assumed our stepper motor circuits would be easy, and never got them working.
• Document as you go. We realized we’d need to power our two stepper motors separately when we designed our power board, but never recorded the reasoning and ended up putting both stepper motors on the same board when we built it. This led to overheating voltage regulators and all sorts of problems.
• Design software for the reality of your sensors, not the ideal. We spent a lot of time developing a wonderful algorithm that assumed perfect sensing, and cut most of it out when our sensors couldn’t live up to those standards.
• Allocate budget ahead of time. If you really want an ultrasonic sensor, make sure you can afford it first. Saying “I’m sure we’ll have enough money for everything else” doesn’t work, because it’s not true. For an ME 218B project, $150 is actually not that much.
• Soldering takes a long time. Budget accordingly.
• Don’t be afraid to cut functionality for the final product. We had to demonstrate both black and green tape sensing for a check-off, but never really felt that green tape sensing was very reliable, so we went back to a black-and-white-only scheme for the grading session. That’s okay, as long as you can solve the big-picture problem that way.
• The tournament is fun, but doesn’t affect your grade. Focus on the requirements for the grading, not the tournament.
• The report for a project of this magnitude takes time to create. Don’t push it off to the last second.
• Organize schedules, budgets, and data sharing with google tool. We used google calendar and google docs a variety of times to share information.
• A 3-5 hour power nap can do wonders for your mental state. You tend to think a lot more clearly, and problems tend get resolved.
• Measure twice cut once. This goes without saying, as most teams only get one shot with the laser cutter. Be prepared to foamcore the things you forget.
• ALWAYS do a continuity check between power and ground before turning on a new or altered circuit.
• When in doubt, check power and ground.
• When in doubt again, check that the E128 is on run.
• When in doubt a third time, check the battery voltage levels.