FTC · For BIOBUZZ · Open source

BIOBUZZ
Robot Simulator.

The team's robot code, plus a browser simulator that runs that exact code on a virtual robot, so setup and tuning can be practised before the robot exists.

16 / 16
Shots scored in the simulator
21
Practice faults built in
9
Guided setup steps
30
Automatic tests
Area
FTC
Built for
FTC Team 19859, for BIOBUZZ
Role
We designed and coded the whole system.
Status
Sim-proven. Real robot pending.
Stack
Java 17 · FTC SDK 12.0 · Pedro Pathing 3.0.1 · Python · JavaScript
What it is

What it
is.

Our team's BIOBUZZ robot is not built yet, but its code is. This project lets us practise anyway: it runs our real robot code on a virtual robot in the browser, so every setup and tuning step can be rehearsed before the real robot exists.

BIOBUZZ is the FTC game in which robots collect game elements and shoot them into goals. Our robot has two shooters, so tuning it by hand is long and easy to get wrong.

How it works

How it
works.

A guided setup, step by step

Nine short programs walk a student through setting up the robot, from checking every motor to tuning each shot, and each one remembers what the last one found.

BIOBUZZ Robot Simulator: the Driver Station with an INIT button and a virtual gamepad on the left, and the Start here checklist on the right with Configure Robot ticked.
The browser Driver Station beside the setup checklist, which ticks itself as each step is done.

Shooting settings that stay reliable

Measured shots are smoothed into a table the robot can trust between the points we actually tested.

The Front Shot Tuner screen: step-by-step instructions and live readings on the left, a field view and a side-on shot camera with a launch angle on the right.
Program 8, the front shot tuner, with the shot camera showing the arc into the CELL.

Practise with problems on purpose

Twenty-one built-in faults, such as a reversed motor, let students learn to spot and fix mistakes before they cost a match.

The result

What it
showed.

With tuning done, both of our autonomous routines score 16 of 16 shots in the simulator and finish with time to spare. That is proven in simulation only; the real robot comes next.

The simulator running the sample blue autonomous: a 0:26 timer and live telemetry on the left, a top view of the field with the robot and the CAD sheet values on the right.
The sample blue auto running on the simulated field, with the CAD sheet the code reads.
What I learned

What I
learned.

Test before the hardware exists

Waiting for the robot would have pushed all of the tuning into the last weeks before competition. Running our real code on a virtual robot first means mistakes show up while they are still cheap to fix.

A long job becomes followable steps

Tuning a two-turret shooter is long and easy to get wrong. Splitting it into nine short programs, each using what the last one found, turned it into something any student on the team can follow.

Proven in the sim is not proven on the field

Sixteen of sixteen shots in the simulator is a strong start, not the finish. Saying clearly that the real robot is still pending is part of reporting the result honestly.