First day, we analyzed the game, making sure everyone understood the game sections, scoring, and robot rules. We also had Homestead, Cupertino, Saint Francis, and Mission San Jose's teams as guests during the brainstorm and game strategy session. One of our alumni, Patrick Wang, gave his analysis of the game's overall strategy, which he presented as an optimization problem.
Initial Decisions
From there, we split into two very large prototyping groups and began working on ball hurdling mechanisms. Following a short team brainstorm session, the two major candidates turned out to be a four-bar arm and an elevator. Due to the bad experiences we had with last year's arm, we are somewhat nervous, but the elevator by nature is far more complex.
Design Review I
After a few days of hardcore prototyping, we held our first design review Thursday night. The electrical team had thrown a Vex robot with omni wheels together, and managed to program it to drive fairly well and demonstrate to the advisers the maneuverability of such a holonomic drivetrain. However, as Mr. Rock, one of our advisers, pointed out, if a four or six wheel conventional drivetrain provides sufficient maneuverability to play the game, then there's really no reason to spend resources developing a holonomic, which we have never done before. There was a toss up between the elevator and arm, as either would work equally well with their own pros and cons, but the end affecter was a virtual consensus.
Meeting together as a team, we recapped what the advisers said at the design review and began our own analysis of the two mechanisms. The arm was by far a simpler design, and thus more reliable and likely lighter than the elevator. However, the massive torque the ball would put on a fairly long arm, combined with our experiences from last year, ended up pushing the team towards the elevator, which the team already has a great deal of design experience with.
Prototyping
For the rest of the week, we worked on getting the base running and fleshing out the designs. Prototyping the end affecter proved the mechanism works in concept, but we still have our misgivings.
Week 2
Week 2
More Prototyping
The design process, is clicking along quite well, and we seem to be making good progress on getting motorized prototypes running. For the elevator, a wooden demonstration of the traditional elevator is done, and the end affecter prototype has been more fleshed out, with a deploying mechanism to allow us to stay in the box.
Design Review II
The advisers again had very valuable feedback, suggesting that we strengthen the grabber by avoiding cantilevering shafts, and adding another fork to the actuating mechanism. Also, asking Mr. Rock about driving straight during hybrid, we settled on trying out the yaw-rate gyro, and possibly the gear tooth sensors.
Fortunately, there doesn't seem to be any major design issue that will prevent us from making our four week deadline. We barely got our base driving in time, managing to miss the deadline by a day, but you could hardly say it drives anyways. The upped drive ratio, combined with the lower coefficient of friction of the 2008 wheels, as well as the total lack of weight on the robot, and the slick concrete we chose to test the base on resulted in it losing traction instantly and managing to not move at all. Quite impressive, if I do say so myself. In either case, even that didn't last very long, as one of the master links wasn't installed correctly and broke. After that, we didn't drive it for a while, since it doesn't matter until we have close to our final weight on the base. Depressing, but at least our robot is driving. We can ship!
Week 3
Week 3
Finishing Prototypes
A wooden prototype elevator seemed rather silly, so the elevator group managed to throw a bosch one together quickly using parts from El Toro 6 and 8. Mr. Shinta was sad to see the spool from El Toro 6 get cannibalized, despite it being his suggestion, but this way we can start our PID testing as soon as possible. The end affecter group is also working away and should be able to finish their design soon. It seems like every day, their blocks get uglier and uglier. At least they don't multiply.
Electrical Issues
On another note, it turns out that everyone's gear tooth sensors are fried, though they seem to be repairable. The IR board also works well on a Vex, and the difference should be neglible on the final robot. Still, it's a mystery how this is all going to work with six remotes on the field at a time. We need to run some interference tests some time soon.
Week 4
Week 4
Running Prototype
Here's crunch time, part one! Mr. Shinta machined parts like mad, and by the end of the week, we were able to get our prototype running. However, we had to resort to wood for the shaft encoder mount. Note for the future, wooden pulleys don't work well with rubber timing belts. All the same, the PID code was tuned quite quickly, and the elevator is rock solid when positioned. That is, with no end affecter, no ball, and four times the gear ratio the final version will have.
Lifting the Ball
After a long time in the woodshop getting riveted, the grabber finally went on the robot, and much to our dismay, not being able to go low enough to pick up the ball. However, propped up on a trashcan, the ball is picked up wonderfully. The PID also seems to hold quite well, despite the weight, and binding doesn't seem to be an issue, yet.
Week 5
Week 5
A Slip?
If we can make a slip anywhere, this is the week. Fortunately, Mr. Appio of De Anza CNC managed to machine the spool for us, so that takes a large load off of Mr. Shinta, but we still have to do quite a few modifications to make the prototypes work properly. Since most of the elevator parts aren't changing, a great deal of those parts have already been finished.
Design Slowdown
The overall pace of design has gotten dangerously slow, likely because the prototype took such a huge load out of everyone. If we can get the end affecter to not interfere with the front wheels and work out the geometries for the pneumatics, though, everything should be fine.
Week 6
Week 6
Finishing Designs
About two weeks late, we finally finished the designs. To both clear the wheels and go low enough, the grabber now fits into the cavity in front of the robot that was made originally for that purpose. However, it weakens the tubing we'll be using to lift the balls in the event that we run into the wall or some other obstacle, though that's a minor issue if we manage to hang on to the ball, or just stay careful.
"Precision" Machining
Fortunately, not too much machining time was lost due to the huge number of parts we have to do this year. Though, now that the rest of the designs are done, a huge influx of drawings went into the pipe. To get everything machined in time, we resorted to making all of the basic structural bars for the elevator and end affecter on the bandsaw, sander, and drill press.
Curse the Breakers, Full Speed Ahead!
With little time to spare, we managed to get the robot together in a somewhat operating state, only to find out that both the grabber and the elevator don't work satisfactorily. We finally got a working configuration on the end affecter forks using copper tubing and solder, but had to double the ratio on the elevator to avoid tripping the breakers, as the motors were drawing over 40 amps rather than the predicted 10.
Second Wind
Though it really looked like we were going to ship a very non-working robot, it turns out Homestead, who is using our facilities as well, since their adviser went on vacation, asked Fed Ex to come later in the day. With the blessed extra time, we threw components onto the robot at light speed, and managed to do a half-hearted full system check out and somewhat hurdle the ball before crating up the robot seconds before the Fed Ex truck arrived. Though we still have a great deal of work to do, we can all sleep fairly well, knowing that the robot works out of the box, batteries not included.
