Skip to content

Stage 2 — Prototyping

Window: Week 1 (late) through end of Week 2. Gate: Prototype Down-Select. Precondition: Strategy Approval passed. Priority List signed.

Prototyping is where we generate data — not where we fall in love with a mechanism. Every prototype must have a written kill criterion before parts are pulled from the shelf.

Core Rules

  1. Parallel tracks. For each priority-1 action, run 2–3 competing mechanism approaches at the same time. For each priority-2 action, at least 1.
  2. Cardboard first. If it cannot be mocked up in cardboard, plywood, or 80/20 in an afternoon, the idea is not concrete enough. Metal comes after data.
  3. Kill criterion in writing before build. Each prototype has a numeric pass/fail written on the whiteboard before the first hole is drilled.
  4. Data ends the debate. The prototype that meets its criterion wins. Not the one with the prettier CAD, not the one the loudest student prefers.
  5. Prototype boards. Each prototype gets its own test rig — motor, motor controller, battery, roboRIO or dev laptop — so software iterates in parallel and prototypes do not compete for the practice-bot electronics.

Prototype Charter (fill before starting)

Prototype Name: (e.g., "Over-the-bumper intake, V1") Owner (student): Mentor advisor: Priority action addressed: (from Priority List) Kill criterion (numeric): (e.g., "acquires 15 of 20 game pieces from floor within 30 s, or dies") Deadline: (e.g., "end of Week 2 Wednesday") Budget: (e.g., "$40 stock, 6 shop hours, 1 NEO, 1 Spark MAX") Data collection method: (e.g., "20 trials, video, count success")

Charter is posted in the shop. No charter, no shop time.

Workflow

1. Enumerate approaches

For each priority action, list mechanism approaches. Example — game piece intake:

  • Over-the-bumper roller intake
  • Under-the-bumper claw
  • Passive funnel + drivetrain-directed

Each approach becomes its own prototype charter.

2. Build in low-cost material

  • Frame: 80/20 extrusion, plywood, or cardboard.
  • Rollers: PVC, foam, compliant wheels off-the-shelf.
  • Actuation: single NEO or Falcon on a Spark MAX / TalonFX.
  • Do not machine plates yet. Do not print custom parts unless a printer is free and it takes < 1 hour.

3. Test against kill criterion

  • Run 20+ trials. Record every trial.
  • Video every session. Review video before deciding pass/fail.
  • Log motor current, RPM, cycle time. Prototype electronics = data-capable electronics.

4. Kill or advance

  • Prototype meets criterion: → advance to Design stage as favored approach.
  • Prototype misses criterion: → killed. Team may propose one revision with tightened scope. Second miss = permanent kill.

5. Down-Select Gate

At end of Week 2, all prototypes present at the Down-Select Gate. Team picks one approach per priority action. See 07_gate_reviews.md.

What to Prototype vs What to Skip

Prototype: - Anything on the Priority List. - Anything with unknown physics (compliant wheel compression, shooter exit velocity, elevator resonance). - Anything mechanically novel to the team.

Skip prototyping: - Drivetrain type if the team has run the same swerve/tank platform for 2+ seasons — trust prior data, order early instead. - Anything COTS-complete (e.g., a MaxSpline gearbox works as advertised; do not re-derive).

Deliverables (Prototype Down-Select Gate)

  • Charter and data sheet for every prototype run.
  • Video evidence.
  • Numeric result vs kill criterion.
  • One-approach-per-priority-action decision signed by lead student + mentor.
  • Rejected approaches documented with reason — feeds engineering notebook and offseason review.

Common Failure Modes

  • No kill criterion. Every prototype "seems to work" and none get killed. Team hits Week 3 with three half-mechanisms.
  • Prototyping in metal. Time and material sunk before data. Kills willingness to abandon a bad idea.
  • One prototype per action. No comparison, no confidence.
  • Prototype uses the same electronics as the practice bot. Serial iteration, halved throughput.
  • Prototype outlives its data. "It missed the criterion but let's keep building." No — kill or write a revised charter.

Handoff to Stage 3 (Design / CAD)

The winning prototype provides: - Working dimensions, motor selection, gear ratio, and compliant-wheel choice. - A physical reference the CAD lead can measure. - Data the CAD gate will use to verify the final design matches proven performance.

Comments