After two-odd weeks spent hibernating in the Packard basement, our team emerged a good deal wiser (and sleep-deprived ;). Certainly, we made mistakes, but at the same time the project gave us invaluable “real-world” experience. Creating and meeting our own deadlines, bringing sketches to physical reality and actually integrating components are parts of the design process that other courses often gloss over or ignore entirely, but the final project gave us the opportunity to meet each aspect head-on.

We briefly summarize some take-home points for the project and advice to future students below:

1. Read the previous reports! There is no point in reinventing the wheel, so to speak. Going over the previous reports is not going to stifle your creativity in designing your project; in fact it is actually more likely to encourage new concepts. Reading the reports, even for different projects, is crucial to understanding what an intense effort the final project is, and how to go about starting the design.
2. Leave room for adjustment. The prime example of this is the shooting mechanism, for which we created a movable track to vary the firing height. This flexible design was invaluable in calibrating three-point accuracy. This was one unusual instance in which changing performance in hardware was much easier than in software. Changing the speed of the gun through code required the overhead of downloading a new version to the robot each time, whereas with the track, we could simply raise or lower the gun as necessary.
3. Stick with one state machine model. If you’re using your own state machine, don’t try to fit it to SES, and vice versa. This can only lead to confusion.
4. Keep backups. Backup code, solid edge models and inexpensive motors in particular. Our shooting motor failed abruptly four hours before competition, much to our dismay. Luckily, our TA, Tenbite, was able to run to the shop and pick up the same model motor in time. As for code & models, the computers in the lab are prone to crashing, so it’s generally a good idea to keep a copy on your AFS account or some analog.
5. Be prepared to redesign. Similar to the second observation, but often redesign is more involved than adjustment. On the first pass, it is very easy to miss calculations, forget spacing, etc. That’s why it’s called a “first-pass” and not a final production! Use the first-pass to identify what areas you may have overlooked, and correct them in your redesign. Redesigning is not the end of the world, but a nonfunctional robot is ;).
6. Keep integration in mind. Though we developed individual parts in parallel, it is definitely recommended to have a solid grasp of the “big picture,” while you make your component. If you don’t keep the big picture in mind, you run the risk of having integration problems.