In which I really don't know what to think of Design Nature.
Design Nature
Design Nature had two parts, the hopper project and the transporter project. The hopper project was individual and all mechanical, and I wrote a lot about it here.
Transporter Project:
The transporter project was rather different. To start, everyone had to write up one or two learning goals for the project. I wanted to learn more CAD and help with some of the machining. Then we each had to propose an animal, the animal's motion, and what the animal would transport. Coming up with a proposal wasn't easy because I wanted to propose an animal no one else had proposed yet (variety seemed like a good idea), and I ended up proposing a walking beaver. After that, everyone had to rank four animals that were proposed, and with those rankings, the professors split us into teams within our studios.
I ended up on team gibbon, and the gibbon was my second choice, I think, behind beaver. How the team's were put together is a little confusing, because as far as animal preference, I know of at least one switch that would have left both people happier, but maybe the professors were also trying to use what they knew about all of us?
What I should have considered earlier on:
I wish I had thought about mechanisms for motions earlier -- while I was thinking about my proposal and when I was ranking animals. Every team had challenges, but some motions are more difficult than others, and I didn't consider that at all when choosing animals. Before the third ModSim project, we each filled out surveys asking how complicated we wanted the physics of our projects to be, and I wish I had thought about something similar here. Gibbons swing by brachiation, which basically involves them acting as a pendulum, free-falling, catching onto the next branch, and repeating. Instead of trying to replicate that, my team tried for a swinging on monkey bars motion, and we didn't manage that, either. I still might have chosen gibbon knowing the mechanism was going to be difficult, but I wish I had known (and had been trying to come up with ideas) from the start.
The other thing I wish I had thought about from the very beginning is the game. The project goal isn't just to build robotic animal toy for fourth graders that transports something; there has to be some kind of game play, as well. It feels like a lot to ask, but at the same time, some teams are really successful in all aspects of the project, so maybe not. My team never had a very strong game idea for the gibbon, though that was partially because we knew that controlling the gibbon's motion would be hard, so we didn't want a complicated game. We let the motion dictate the game instead of leaving both open in our ideation phases. In the first meeting, my team decided our gibbon was going to swing in this particular way, and then we started coming up with game ideas. (We also never considered that the gibbon didn't *have* to be robotic -- one of the successful teams in our studio did an owl that the kids swung around on a string. It had no moving parts.)
What I wish my team and I had done differently:
Through the learning goals and some discussions in the auditorium, the instructors made us clear that they wanted us doing types of work with which we weren't comfortable yet. On Team Gibbon, at least, this resulted in people being in roles that had them way out of their depth. For example, I was the leader of the CAD subteam because I had written that I wanted to learn and practice more CAD. I think I should have been assigned the CAD for some major subsystem, like the body (which I did on the first prototype), but I definitely didn't know enough to be leading the team. I didn't think about the fact that everyone should specify material in their CAD, add reference geometry, or make sure everything was fully defined -- I didn't really know what those last two were until the second prototype. I made (and didn't catch) so many mistakes in my own CAD that I really shouldn't have been responsible for checking other people's. After the first prototype, we rearranged roles but never clearly defined again what each person should have been doing. I think the better reaction would have been to have people who knew what they were doing as sub-team leaders without having them do much of that particular work. (Admittedly, that makes more sense with things like machining and CAD than with project management.)
I wish I had done more work on the second and third prototypes. I helped with some CAD and the assembly of the second prototype, but on the third my major contribution was the head and the fur. I didn't spend as much time in studio working as I feel like I should have. However, one of the two people on the team who did the most work (and spent absurd numbers of hours in studio) told me that in general he felt like I was there, even if I was just filing a piece down or holding something steady as he drilled, and sometimes it was being there for the long hours of work that mattered.
What I wish had been different in my studio:
I think Design Nature would have gone better if class time were a little more like class time in ModSim. The professors did go around talking to each team, and in a survey of the first years, we found that there was a significant difference in DN satisfaction based on studio professor, and my guess is that this has to do with the in-class feedback. My team felt like every time the professor talked to us, we were told all kinds of things that were wrong, directions that we weren't going that we should be going, but we got no encouragement or guidance. After a NINJA or instructor visit in ModSim, I felt better about the project; after an instructor visit in DN I always felt worse. What I've heard about the professor in the most satisfied studio is that he gave really helpful feedback as he circulated.
Overall:
I disliked Design Nature, and part of the way through the semester I realized that it doesn't seem like it has changed much over the years. The only change, as far as I can tell, is in the final project: it started with wall climbers, then changed to swimmers for some reason, and then having to deal with water was a pain, so the project became transporters. Most classes change often at Olin, but Design Nature is one of the few classes where people several years apart can have nearly identical experiences. I was talking about this with some sophomores, though, and one of them said, "They don't change it because it works."
The goal of Design Nature is to teach the design process, like ModSim teaches the modeling process. And yes, I do know what the design process is now. I've gone through it twice (neither time with much success). I know to start with the values and needs of the users to get design requirements, then to ideate, sketch and sketch model a few ideas, do an initial prototype, and loop back through. I know to work with a few ideas in parallel and not narrow too early. I know to always keep the values and needs in mind. I suppose, because I know these things, that I learned what I was supposed to learn from Design Nature. But while some of this was reinforced by mistakes I or my team made, I think I know these things because I was told them -- in Design Nature auditorium lectures, in the order in which assignments were due for the hopper project, in the transporter design review evaluation sheets.
I learned what I was supposed to learn, so maybe the class worked. That doesn't mean it worked as well as it could have. That doesn't mean it couldn't change and become better.
No comments:
Post a Comment