When the 2015 game was announced in mid-September, our rookie team shifted into high gear brainstorming to begin the process of designing a robot to perform the appointed tasks. We knew that the big challenges would include:
- Climbing the steep, slick, and obstacle-loaded mountain
- Moving the ball and block debris
- Finding the rescue beacon and pressing the right button based on color
Our earliest efforts used the “Pushbot,” an example robot design provided by FTC using Tetrix parts. We also acquired our own playing field for testing, buying a set of floor mats and a half-set of game elements, and building our own surrounding fence from PVC pipe and plywood.
It became clear from the outset that climbing the mountain would be essentially impossible with wheels and probably difficult with treads. Also, the challenge of autonomously finding and pushing the button on the rescue beacon required a lot of knowledge about sensors and programming that we lacked.
After several wheel variations were tried, we searched the internet for a good tractor tread system, settling on the Lynxmotion tread, featuring molded plastic pads with a soft rubber surface.
These worked well for us, but caused some panic when we tested our first implementation and discovered that they were capable of digging grooves in the foam rubber floor of the playing field. This is a rule violation and would disqualify our robot if tested during competition. The traction/floor damage test is done by powering the robot full speed against an immovable object, spinning the treads against the floor for 30 seconds. Fortunately, we learned that the problem with our first model was caused by too much force at the rear end of the tread caused by torque from the front of the robot being higher than the rear. Shifting the point of contact with the immovable object closer to the floor eliminated the issue.
When we tried to make our robot with treads climb the mountain, it worked fine until encountering the first churro bar. These are aluminum extruded bars that extend across the artificial mountain path about 1.5″ above the surface of it.
Without some other aid, the treads would not climb up over the bar, even when we raised the front of the tread like a World War One tank:
We had planned from the outset to attend the qualifying tournament at Smithville to observe but not to compete. The date of this event was December 5, 2015 and we were nowhere near having a functional robot. This event was very useful to us, however, and we learned a great many things that caused us to redirect and refocus our design effort.
First and very surprising to us, was that very few teams attempted any autonomous behavior and very few teams had sensors in use on their robots. Second, many robots had their plan disrupted by debris. The balls and blocks would get in the way of their desired paths, or worse, get stuck under the robots and prevent them from moving at all. It looked as though many teams practiced without debris.
There was one team, #7357 from Lee’s Summit, that performed spectacularly well and had a huge influence on our thinking. Their robot was very sophisticated with over 30 custom 3-D printed parts. In autonomous mode, it carefully swept debris from in front of the mountain, turned toward the path, extended a tape arm, and pulled up to the high zone. It did this very reliably. Then in operator-controlled mode, it backed down the mountain, dumped climbers into the bucket behind the rescue beacon, climbed the mountain again, flipping the side toggles to release the zip line climbers, and then in the end game, pulled the all clear signal and hung from the pull-up bar.
Their tape arm, a 3-D printed design using a carpenter’s steel tape, was particularly interesting because it overcame all issues of climbing the mountain and allowed hanging at the end.
Our Strategic Spreadsheet
We put together a spreadsheet, comparing the points available for various tasks with the perceived difficulty of accomplishing them. The perception of difficulty changed after our Smithville Qualifier experience. Using a difficulty scale of 1 to 5 (5 being the hardest), we calculated Points/Difficulty and ranked this number in three groups: P/D <= 5, P/D >=10, and in between:
Our strategy then focused on these elements:
- Ignore the Autonomous period.
- Design and build a tape arm.
- Use Driver-Controlled mode to climb the mountain and trip the Zip Line climbers.
- Hang from the Cliff Pull-up Bar in the End Game.