Hi, I'm Lewis!

Mechanical Prototyping

Weeks 1 to 4, and still going

Learn how the team turns a concept into a working mechanism by prototyping the robot intake, from the first belt idea to a tested compression system.

• Built and improved the robot belt prototype. After fixing frame flex the intake moved 8 balls per second in testing. A video of it running is linked below.
• Designed the gripper prototype, then built intake versions with bearing retention and bumper clearance fixes.
• Assembled the intake across several weeks and set its compression by testing. It went from loose and inconsistent to a repeatable grip.
• Ran intake checks alongside the CAD rotation in the last week of the quarter.

Video of the belt prototype running ↗

Prototypes fail on measurement, not on ideas. Belt tension, bearing fit and compression distance all needed numbers before the design came together.

CNC and Fusion 360

Week 3

Learn CNC workflow and Fusion 360 well enough to go from a model to a cut part.

• Practiced CAM setup and toolpaths in Fusion 360, then cut parts on the CNC.
• Hit a CNC setup problem mid week, diagnosed it instead of dropping the machine, and moved to cutting time once it was stable.

CAM is where designs get tested against reality. Tolerances and workholding decide whether the part that comes out matches the model.

Electrical

Week 4

Learn the robot electrical system, CAN bus wiring, and how to trace faults.

• Wired and checked CAN bus chains with the team.
• Learned fault isolation. When the root cause was not obvious I tested one suspect component at a time instead of swapping parts at random, which found a wiring fault faster.

Electrical problems are process problems. A sequence beats intuition when a chain of devices drops off the bus.

Code

Week 5

Learn the robot software stack, WPILib, PID control, logging, and the team Git workflow.

• Implemented PID rotation control and explained why closed loop control differs from open loop on our Kraken x60 drive with its 14 to 81 gear reduction and 3 volt command cap.
• Set up DogLog tunables so control gains can change without a redeploy, and worked through the team Git workflow for shared code.

Tuning is measurement. Log first, then adjust. That is what makes PID settle instead of wander.

Machining

Week 6

Learn the shop machines, the drill press, the lathe and the horizontal band saw, and the accuracy habits that go with them.

• Used a center finder on the mill and read its oscillation to square up work.
• Used the paper shim method to hit a snug press fit. Both tricks work because they turn an invisible size into a visible reading.
• Machined intake parts to print dimensions for the assembly push that week.

Precision is a habit of checking, not a property of the machine. Every measured cut start beats every guessed one.

Assembly

Week 7

Learn how the team assembles a drivetrain, alignment methods, torque practice and hardware choices.

• Rebuilt a swerve drive module with the team, from frame up.
• Learned why assembly order matters. Tighten the chain of fits in sequence rather than each joint in random order.

A rebuild is where design mistakes surface. Assembly feedback went straight back into the intake and drive base designs.

CAD in Onshape

Week 8

Learn Onshape well enough to model real robot structure from shop measurements.

• Modeled the drive base. Four 1/8 inch side bars framing a square chassis with an imported swerve module at each corner.
• Hit a part alignment problem in the assembly, fixed it with planar and parallel mates after asking teammates, and finished the base CAD.

CAD is the handshake between design and build. If the model is wrong or misaligned, every shop downstream pays for it.

Beyond the rotations

Across the quarter

• Two threads ran alongside the seven rotations. Driver tryouts, where I ran cone circuits and figure 8 drills, and helped pull robot logs over SCP when the robot kept disabling itself outside. And team onboarding in week 1, where the group set norms and I picked up the habit of logging work the same day it happens.

Drive Train Test Board

Robotics 1

The goal was to build a model drive train on a test board and get familiar with building drive trains for future robots. We needed a system that could turn a wheel with a motor, sprockets and a chain. Along the way I learned Fusion 360, the lathe and the drill press.

• Start. We started where another group left off, with just a motor and one sprocket.
• Sprocket shaft designs. We designed sprocket shafts in Fusion 360 from scratch with our own measurements. This was the final design after a few iterations, because of trouble with the program and some measuring mistakes.
• Chain. We cut a chain to a workable length, then held another gear in place to mark where the hole for the second sprocket needed to go.
• Cutting the hole. We used a drill press to cut the hole using our pencil marks.
• Sprocket shaft. We used a lathe and followed our designs to cut the sprocket shafts to the size we needed.
• Assembling. We assembled the bearings, sprockets and shafts, then put the chain on.
• Finished product. The finished drive train worked on the first test. The project went smoothly overall, besides a few issues and mistakes in the design phase.

Minibot

Robotics 1

An obstacle course tested different objectives, with points based on difficulty. The goal was to build a good robot and drive it well enough to score the most points.

• Define. Our group had to build a working minibot. We finished on schedule, with time left to test before the course. We split the work so we could build together without getting in each other's way.
• Build. We built an 8 by 24 inch wooden platform to hold the mechanical and electronic parts, drilled the holes for the breaker and power distribution hub, planned where every piece of electronics would sit, and drilled larger holes for the gearbox motor and the side board screws.
• Test and iterate. We used the extra time to test outside on the course. Every small issue, like one wheel not touching the ground or uneven power to the motors, got fixed so the bot could finish the course.
• Implement. In the final run the robot did exactly what it did in practice, which was very well. What we missed was the battery. We had used too much charge during practice, so the robot ran weaker and could not finish the course.