Here are some competition photos along with the changes. Photo credit goes to Dan Ernst.
The drivetrain, tower supports, and arm were not iterated during the competition season with the exception of two parts that developed stress fractures. We had to replace two window motor couplers on our shoulder joint in St. Louis. We felt that aside from those two parts, these subsystems performed wonderfully, so we left them how they were.
End effector changes:
As I mentioned when I first revealed the robot, we smashed our end effector. We originally used 1/16th aluminum angle brackets to make it as light as possible, but it wasn’t very strong. The combination of lack of strength and tweaks to our operating software are the reason we smashed it. We rebuilt it with 1/8th aluminum angle brackets and also added some gussets to increase strength. We also opened up the opening of the claw to create a funnel shape.
Deployment:
We didn’t do an incredible amount to the deployment system, but we did enough to improve it between Waterford and St. Louis. We had to replace a drawer slide between Waterford and Ann Arbor and for some reason, the new slide had more slop than the original, so we would occasionally mis-align. To fix this problem, we added a lexan fishtail to align ourselves when we deployed. After the Michigan State Championship, we realized that if we were to continue to be competitive, we needed to speed up our deployment system. We simply bought a lead screw with coarser threads and cut our deployment from 1.25 seconds to .6 seconds.
(before)
(after…a bit harder to see though)
Minibot:
Our original minibot(seen in photo 2) was made with almost all tetrix parts. It was slow, climbing the pole in around 4 seconds, but it was reliable. At Waterford, it was the reason why we were first seed - it climbed the pole 14/16 times. Like our deployment system, we realized that to be competitive with the minibot, we needed to speed up. We switched designs to the typical lightweight, direct drive, speed demon minibot(seen in photo 3). The chassis was a piece of lexan bent in a channel shape. A piece of tetrix angle was riveted to the lexan to mount the motors. Both motors were taped to the angle bracket and powered one shaft. An idler shaft was sat on the other side of two magnets. A tripwire attached to the hostbot tripped a limit switch to turn it on and a sliding piece of lexan slid another limit switch into a bolt head to shut the minibot off when we hit the top. We visited 2081’s pit in St. Louis and they clocked us at 1.129 seconds, and yes, it did win us races.
http://a7.sphotos.ak.fbcdn.net/hphotos-ak-ash4/222350_2025520919943_1301636643_2494895_5022944_n.jpg
We didn’t win any blue banners, but we built a fantastic robot that won the Industrial Design award at Waterford. Velociraptor did everything we wanted it to do with great ease. We did quite a bit more iterating this year than in previous years and it showed us another part of the engineering process. We are planning on attending the Tardec Invitational at IGVC and IRI. If you happen to be at either competition, please stop by and have a closer look.
