WEEK 3 UPDATES
Designed in-house by 7407 students, BoarBot is an affordable, accessible, and competitive robot designed for rookie FRC teams. For its design, we’ve decided upon a nimble robot combined with the main philosophy of “touch it, own it.” The ability to completely control a game piece at every angle quickly will be extremely beneficial. BoarBot’s goal is to score only in low because we established that the ability to score at every level was too lofty a goal for BoarBot’s mission. We established that a roller claw would be the most efficient way to intake game pieces. This would also shorten the wheelbase to ensure room for other wide robots on the Charge Station. In fact, the robot as a whole is relatively small in order to do just that. Last week, we finished the the CAD to the point where we felt comfortable to start manufacturing. As we began manufacturing, we ran into some issues and had to rework the CAD, notably with the size of the claw, which we had to increase. We also analyzed the advantages and disadvantages of Everybot, a similarly-missioned robot to which we’ve taken a totally different route from. While Everybot is aiming to score high and intake only from the top station, BoarBot will be able to intake all its pieces from the ground. Everybot might be too skill-demanding for students to build and potentially too expensive for rookie teams (Everybot’s bill of materials costs ~$1700 compared to BoarBot’s ~$700 cost.) We cannot rely on these teams to have full machine shops, so BoarBot can be constructed with more widely available tools. BoarBot is coming along nicely and we will share more updates with it as it develops.
This pneumatic intake/extake prototype was designed with inspiration from Ri3D Redux’s testing. The intake is pretty simple in concept; it compresses the cone/cube from both sides to hold it using a pneumatic. This prototype allows the game piece to swivel under gravity as the pads holding the piece are attached to bearings. The intake worked very well for cubes but had a couple of issues with cones. First, if we grabbed the cone too low, gravity would swivel the top of the cone down, making it entirely impossible to score. A second issue we noticed was that if the cone was grabbed too far off-center, it would not be perfectly vertical. A solution to this second issue could be making the grabbing pads larger. Our main takeaway from this prototype is that this idea would work, but it would require a high level of precision in many ways. For that reason, we are putting this grabber concept on the back burner while we look at possible ways to make a grabber with rollers. Assembly has begun on a wrist-and-claw system.
Intake/Extake Cone Video
Intake/Extake Cube Video
Our climber is meant to fix the issue of three robots not being able to easily fit on the Charge Station. Assisted climb enables a last second engaged climb onto an already engaged alliance partner. This concept makes the climb fast, which means that time normally spent trying to get three robots to coordinate a balance can now be spent scoring additional game pieces. Additionally, during alliance selection, we can choose large robots as our alliance partners. There are two big points to our assisted climb concept: 1) the robot is contacting only the Charge Station and an alliance partner fully supported by the charge station, and 2) being below the Charge Station lowers the station’s center of gravity, making it easier to balance. We use elastic to keep the two parts of the climb together during deployment and stowage. This elastic is overcome when the robot climbs. Also, there is a ratchet on the spool so that when the robot powers down at the end of the match, we stay climbed. The ratchet’s pawl is located on a pneumatic cylinder, so we can abort a climb if anything goes horribly wrong. Carbon fiber rods are used for the forks because they are lightweight and really strong. The winch uses a 16:1 gear ratio for a MAXPlanetary gearbox with NEO motors and lathe-made spools. The climb sequence and photos of the climber CAD can be seen in slides 13-19 of this presentation that we used in our appearance on the OA Show (which you should check out if you haven’t already).
The assembly sub-team wrapped up in drivetrain assembly early this week. We worked on electrical finalizing placement on the minicomputer, pigeon, PDP, radio, pneumatics, etc. With these finalized we could work with manufacturing to make polycarbonate brackets and covers for the swerve pods, Spark Maxes, and battery. Since handing the drivetrain over to programming, assembly has been focused on the claw and the elevator. We took some basic first steps with the claw, but we are currently waiting on the rest of the hardware needed (compact tie rod air cylinder among other things). With the elevator, we have been focused on finishing the installation of the Thrifty bearing blocks. We finished painting and worked on the main elevator assembly using our custom MarkForged tube inserts. We also finished manufacturing our plates for the elevator and attached the gears/sprockets for our chain and pulley system. The final steps are lightening on the 2x2 for the last stage of the elevator and then shifting to the wrist, claw, and intake assembly.
Over the weekend and this week, the electrical team had been hard at work wiring the drivetrain. While waiting for the design team to finish the CAD, we decided to wire everything we could. This included the SparkMaxes, Cancoders, and the other necessary components. In order to make our electrical system cleaner this year, we also decided to make swerve covers to mount the controllers, and grommets in our crossbar to make the entire wiring process cleaner. We also started making a robot map, which will make programming and electrical maintenance much easier during the season.
This week, the programming team focused on setting up and making swerve code for our new drivetrain and incorporating Photonvision april-tag detection into its field odometry. We also worked on making skeleton code for subsystems that are currently being manufactured, like the elevator and claw. The new wpimath update set us back a bit because we are having some issues regarding the new swerve module position-based odometry, but we’re hoping to resolve those issues soon.
Week 3 Swerve Drive Video
Again, the link to our full robot CAD is in the first post for those who want to see more. And as always, don’t hesitate to reach out with any questions or clarifications!