For the FIRST Steamworks challenge, the team endeavored to design, build, program, and operate a robot that would pick up the gears and deliver them to the airship, enabling it to be prepared for takeoff. The robot was built around a central purpose of climbing the rope at the end of the match and being able to actively acquire gears from either the floor or the distribution box. The team considered many drive systems, finally settling on a hybrid of mecanum and omni-directional as it was the most maneuverable and robust for what they expected the game would require.
This year’s robot is focused on picking up the gears and climbing the rope. Careful evaluation of the rules led the team to forego ball handling in favor of more valuable points in placing the gears and climbing the rope at the end of the match.
A chain driven spindle is designed to engage with the knot on our rope. The “bunny ears” corral the rope, guiding it to the center, where it quickly wraps around the spindle, pulling the robot up as it is collected.
The robot uses a hybid holonomic drive with two mecanum wheels in the front and two omni-weels in the back. To facilitate high control and maneuverability, the mecanum wheels are oriented 90 degrees from their normal orientation.
To get clearance under the robot for gear handling, we created transfer cases that would connect mini-CIMs and Versa Planetary Gearboxes to the mecanum and omni-wheels using internal belts and pulleys.
We planned the layout of the electrical system to utilize powerpole panel mounts as pass-throughs from the top to the underside of the robot. A custom circuit LED light array was installed to illuminate the vision targets on the field.
WIth the airship and hundreds of yellow wiffle balls obstructing the drivers’ views of the field, the ability to use vision processing to guide the robot towards gears, the elevator, and the rope will make driving more efficient.
In addition to using LabView, we are using Python this year for vision processing on the Jetson TK-1. The TK-1 will communicate with the RoboRio so that our code, as a whole, runs more efficiently for better control of the robot.
The challenge in picking up the gears is thair lack of grippable profile. To combat this feature, we created two gear engagement modules - a conveyor system and a claw system. With a universal mounting system underneath the robot, both systems are under simulatneous development for use at competition.