A Closer look at the changes made for the mechanical side and thoughts for the future
PoE teams Assigned 2/28 Decided to order equipment 3/11/13 Spring Break 3/18/13 – hardware arrived this week. Did not end up scheduling team meetings because 1 team member was on vacation. Design Review 1 3/28/13 – demonstrated first prototype; realized current technique of bending sintra after drilling holes would not work for mass production as it was hard to get holes to line up. Also realized that the tension on the tracks and the difficulty in attaching the hubs to the shaft of the wheels meant it would not be straightforward to assemble a tracked vehicle. Track idea scrapped (after purchasing 5 sets). Plans to design a wheel that would directly attach to the drive shaft shape put in place. Waiting on Electrical hardware in order to determine desired size of robot. Design Review 2 4/18/13– the weeks between this and the last design review were spent designing and fabricating more sketch-model level prototypes in order to gauge the correct size we’d need for all of the components. First pass of lasercut parts done 4/22/13– we manufactured bot out of acrylic instead of foamcore. Bot worked but was hard to put together and was more flexible than desired, so we made plans for another revision. Additionally, we wanted to engrave “Olin SwarmBots” on our robot, which turned out to be much more difficult than anticipated (changing line thickness/color of text is not trivial in Solidworks). 2nd and 3rd pass of lasercut parts done 4/29/13 – manufactured two different designs to see if one would work better than the other. Spring loaded iteration worked well but still needed glue to hold in place. Crossbraces allowed for easier gluing of pieces at 90 degree angles. Much stiffer box. Robots assembled 4/30/13. Used metal L-brackets lying around in PoE room and balsa wood and cardstock to attach mouse encoder to bottom of robot. Even at larger scale, difficult to grab parts and pieces inside body of bot. Decided to mount electronics on top.
Future considerations: There are some changes that should probably happen to the CAD to make them easier to assemble in order to put together robots without gluing them together, but that will be left up to the next group who will work on this.
Recommendations for future teams:
Making sketch models is great, but having set deadlines to fabricate the current CAD on a certain day with acyrlic would probably have sped up the prototyping process even more. In retrospect making a robot every week may have been beneficial. If we had made the acyrlic pieces sooner we could have had a driving robot earlier. (Thickness of foamcore did not allow for wheels to attach to drive shafts.) May have just been better to make a robot that was definitely oversized for our electronics instead of waiting for the boards to be designed and fabricated.
Design, Fabricate, Evaluate. Repeat as many times as possible. Even if you don’t have the resources to make parts from nicer material, find some way to create a cheap prototype. It’ll help give you a better perspective on size and how everything would be integrated. Modularize. Always design knowing that you wil have to replace motors at some point. If you’re making things in quantity, make sure there is an efficient process to manufacturing parts quickly and assembling your system efficiently. The first one can be slow and give great insights into what you should do to make the process faster, but after building and assembling your own prototypes, even building just five can get tiring.