Skimmer

Architecture at a glance

Dual-feed concept — the architecture's defining feature

Skimmer's polycarb tube has two distinct feeding paths — this is what makes the architecture interesting and what justifies the build complexity. Each path requires the swing arm to be in a different rest position before the element arrives:

1. Path A (front flex wheels → tube, swing arm at Position A): for pins or cups lying flat on the field. The driver positions the swing arm at the forward-down rest (Position A) so the tube hovers just above the front flex wheels with mouth angled toward the flex wheels. The driver then drives forward over the element; the front flex wheels pick it up and convey it upward into the tube's open mouth.
2. Path B (loader drop → tube, swing arm at Position B): for cups and cup-on-pin combos delivered from the alliance loader. The driver positions the swing arm at the vertical rest (Position B, arm at +90°) so the tube is vertical with mouth pointing straight up. The driver then positions the robot back-end-first under the alliance loader; the human player drops the element from above and it falls into the tube via gravity. The pneumatic cinch closes around it.

Both paths converge into the same tube manipulator. Once the element is captured, the swing arm sweeps to the appropriate angle to deposit at any of the three goal heights. Note that elements lying on the mats are typically horizontal (the game manual's pin spec is 6.5″ long, 1.6″ hex; cups are also wider than tall when they fall on their side) — so the front flex wheels are sized to grab from above a flat element, not to scoop a vertical one.

Side view — tower, two rest positions, and the sweep between them

The swing arm pivots from the top of a tower mounted at the chassis geometric center — not at chassis back as in earlier revisions of this page. The center-tower placement is mandatory for SG2 compliance: with a 14″ arm and tower at chassis back, swinging the arm to horizontal-back (the loader-catch position) would put the arm tip 12.7″ past chassis back = 30.7″ total horizontal expansion, exceeding the SG2 24″ limit by nearly 7″. Moving the tower to chassis center and shortening the arm + tube to 12″ total reach lets both rest positions sit exactly at the SG2 envelope without violating it.

Skimmer's swing arm has two functional rest positions, not one, because elements arrive from two different directions and need two different tube poses to catch:

• Position A — intake handoff (arm forward-down at −50°): the tube angled forward-and-down toward the chassis-front flex-wheel array, tube mouth at ~2.83″ off tile and ~1.25″ inside chassis from front edge. The flex wheels lift elements off the mat and deliver them into the tube mouth from below.
• Position B — loader catch (arm horizontal back at +180°): the swing arm rotates all the way around through vertical-up and continues to horizontal-back, positioning the tube at the back of the robot, tube mouth at 12″ off tile and 3″ past chassis back edge — just above the SG11-raised loader chute exit at 11.85″ off tile. The human player drops elements into the loader; the loader's bottom exit delivers them to the tube. This rest is opposite the flex-wheel intake, at the chassis-back side as Coach T specified.

Goal scoring happens during the sweep between these two rest positions — the arm pauses at an intermediate angle with the tube tipped to release into a goal at the appropriate height. The sweep is approximately 230° because the two rest positions are on opposite sides of the pivot (forward-down vs. horizontal-back), and the arm goes UP and OVER through vertical to get between them. This is a long sweep but mechanically feasible with the 11W Red 100 RPM swing-arm motor (one full 230° transit takes ~0.4 s at the motor's no-load speed; loaded it's slower).

The two-rest design — why it works

Most swing-arm scorers have one rest (mouth-up at the loader) and one extension (mouth-forward at the goal). Skimmer needs two rests because elements arrive from two different directions: from below via the front flex wheels, and from above via the loader. The tube has to be in different orientations to receive each. A single rest position can't serve both.

• The 90° arc between rests is small enough that the swing arm motor doesn't spend most of the match in motion — it parks at one rest, intakes, sweeps to score, returns to the appropriate rest based on what's next.
• Position A's tube orientation (horizontal, mouth toward flex wheels) means the front flex wheels' lifting motion deposits elements directly into the tube's open end — no "throw and hope" handoff.
• Position B's tube orientation (vertical, mouth up) means the loader human player can drop elements at any angle of approach and they fall straight in via gravity.
• Cups and pins lying on the mats are typically horizontal (the game manual's pin spec is 6.5″ long, 1.6″ hex section — they roll and lie flat). The front flex wheels are sized to grab a horizontal element from above; this is why Skimmer's flex wheels are at floor level and Path A only works for elements lying down.

The tube-mounting question — OPEN, needs bench validation

A 2-bar swing arm has a known kinematic property: the end-effector orientation rotates 1:1 with the arm angle (a 230° sweep flips the end-effector by 230°). For Skimmer this gives three tube-mounting options, each with trade-offs:

• Tube along arm (what the current SVG shows): mouth points outward from pivot along arm direction. At Position A (−50°), mouth points forward-and-down toward the flex wheels — good for flex-wheel handoff. At Position B (+180° horizontal-back), mouth points horizontal-back — bad for catching a vertical drop. Fix: add a 3D-printed funnel cap at the mouth that catches vertically-falling elements and redirects them sideways into the horizontal tube. Funnel adds ~1.5″ of effective tube length; verify SG2 still passes with funnel.
• Tube perpendicular to arm (cross-axis): mouth always points one direction relative to the arm. If mounted so mouth points UP at Position B (when arm is horizontal-back, perpendicular = up), the same mounting means mouth points horizontally at Position A — bad for flex-wheel handoff. Fix: redesign Position A to have arm angled differently so the perpendicular tube mouth points toward the flex wheels.
• Tube on a passive pivot (wrist mechanism): tube rotates freely at arm tip, gravity keeps mouth UP regardless of arm angle. Catches loader drops cleanly at Position B; mouth flips to face flex wheels at Position A. Cost: adds mechanical complexity (one more bearing + counterweight or spring) and may need a stop to prevent tube swinging during sweep transit.

The 5.5W tube rotation motor (around tube's own long axis) is independent of this mounting decision — it handles in-cycle rotation for cup-retention fine-tuning regardless of which mounting is chosen. Bench-test all three options with a cardboard mock-up of the loader chute at 11.85″ off tile and document the chosen tube-mounting axis in EN4 with photos and the rejection rationale for the other two. This is exactly the kind of "show your work" geometry decision the engineering notebook rewards.

Design iteration log — tower position revision (2026-05-10)

1. What changed (TL;DR)

The Skimmer geometry was revised on 2026-05-10 after Coach T flagged that the v1 SVG appeared to place the swing-arm tower too far back, which would push the arm past the SG2 24″ horizontal expansion limit when swung toward the loader. A constraint review confirmed two independent problems with the v1 geometry:

• The v1 Position B tube mouth was at 30″ off field tile (arm vertical-up + tube along arm). The 2026-27 Appendix A loader spec released 2026-04-27 puts the chute exit at 11.85″ off tile in the SG11-raised state (or 3.25″ lowered). A 30″-high tube mouth could never catch an element dropping from an 11.85″ chute exit; the geometries were mismatched by ~18″.
• If we corrected Position B to actually reach the 11.85″ chute exit, the arm would have to swing past vertical-up to horizontal-back. With the v1 tower 1.33″ from chassis back and a 14″ arm, horizontal-back position puts the arm tip 12.67″ past chassis back edge — total horizontal expansion 30.67″, violating SG2 by 6.67″.

The fix moves the tower from chassis-back to chassis-center and shortens the arm + tube to 12″ total reach (8″ arm + 4″ tube along it). With the pivot at chassis center, the arm can swing to horizontal-back with the tube mouth landing exactly 3″ past chassis back edge at 12″ off tile — precisely at the SG2 envelope, just above the 11.85″ chute exit.

2. v1 vs v2 geometry comparison

3. The constraint math (showing the SG2 violation)

The v2 reach of 12″ is precisely the maximum the chassis-center pivot allows under SG2. Any longer total reach would put one or both Position B mouth coordinates past the 3″ chassis-edge expansion limit. Any shorter would leave usable SG2 margin but at the cost of not reaching the 11.85″ chute exit position past chassis back. 12″ total reach is geometrically the only correct answer.

4. Trade-offs introduced by v2 (and why we accepted them)

1. Sweep arc grew from 120° to 230°. The motor must rotate the arm through nearly a full revolution between rest positions, rather than the v1 quarter-turn. At the 11W Red 100 RPM swing-arm motor's no-load speed, the v1 120° sweep takes ~0.2 s; the v2 230° sweep takes ~0.38 s. Under load (arm + tube + element), realistic transit time is 1.0 to 1.5 s. Why accepted: the swept positions are visited only at the start/end of each cycle, not continuously, so the additional 0.2 s transit cost is offset against being able to play the game at all (the v1 geometry can't legally compete).
2. Position A mouth moved from “just past chassis front” to “1.25″ inside chassis from front edge.” The shorter v2 reach pulls the mouth inside the chassis envelope rather than projecting it out beyond the flex-wheel array. Why accepted: the flex-wheel array still does the active intake work (pulling elements off the field into the chassis), and the tube mouth at 1.25″ inside chassis from front edge is still inside the handoff zone. The wheels carry the element 1-2″ further into the chassis than they would have under v1; the difference is small and well within the handoff geometry the team will tune during bench testing.
3. Position B mouth orientation changed from vertical-up to horizontal-back. In v1, vertical-up arm gave a vertical tube with mouth opening up — perfect for catching vertically-falling elements from the loader chute. In v2, horizontal-back arm gives a horizontal tube with mouth opening back — bad for catching a vertical drop, since the element would hit the side of the tube rather than enter the mouth. Why accepted (provisionally): this is the only geometric option that satisfies SG2 AND reaches the chute exit position. The mouth orientation problem is treated as a separate sub-problem (the “tube-mounting question” section above) with three candidate solutions for the team to bench-test: funnel cap, perpendicular tube mount, or passive-pivot wrist. The team should NOT proceed to final fabrication until one of these three is validated against an actual loader drop.
4. Tower at chassis center may interfere with battery/brain placement. The v1 chassis-back tower left chassis center clear for batteries, brain, and wiring. v2 puts a 2″ × 2″ tower footprint at the geometric center of the chassis, which is normally prime real estate for electronics. Why accepted: the tower footprint is small (4 sq.in.) and can be worked around with creative electronics mounting (e.g., battery on a side plate, brain near a chassis corner). The team should plan electronics layout around the central tower before drilling chassis mounts.

5. Suggested EN4 entry text (the team can adapt this)

6. Lessons (for the team's future iterations)

1. Cross-check geometry against the game manual at EVERY arm angle, not just at rest positions. v1 passed SG2 at the two rest positions drawn but the actual loader-catch position the team would have needed in match play would have violated SG2 by ~10″. Future designs should sweep the arm through a software check (or paper-and-pencil sweep table) confirming compliance across the entire angle range, not just at "the position we'll usually be in."
2. Numbers from the game manual are not interchangeable with engineering intuition. The team's intuition that “vertical tube catches vertical drop from loader” was geometrically correct but missed the actual chute exit height (11.85″), making the geometry produce a mouth at 30″ off tile catching from nothing. Always pull the specific dimensions from the published spec before computing reach math.
3. Two-position rest geometries with very different reach requirements may not fit a single fixed-length arm. Position A and Position B target points at very different distances from a chassis-back pivot. The single-length-arm constraint forced us to compromise on Position A (mouth now inside chassis rather than projecting forward). For future designs with similar dual-position requirements, consider compound mechanisms (wrist, telescoping arm, parallelogram offset) before committing to a simple swing arm.
4. Solving one geometric constraint often creates a new sub-problem. Fixing the SG2 violation introduced the horizontal-back mouth orientation problem. Design iteration is not a sequence of completed steps; it’s a sequence of newly-visible-problems. Plan for sub-problems to appear and budget bench time for them.
5. Document rejected designs as carefully as accepted ones. The v1 geometry is now superseded but the analysis that proved it infeasible is valuable engineering work. The notebook should retain v1's specs, the SG2 violation math, and the rejection rationale — this is what judges score, not just the polished final design.

Motor budget — 82.5W or 88W (depends on intake motor choice)

The 5.5 W spare is real headroom — it can become a 5.5 W half-motor 6th drive (slight torque/redundancy boost) or stay as a sensor port. Skimmer is the only fleet bot besides Heron with budget slack; Falcon and original Mantis are at the cap.

Torque analysis — swing arm load sizing

Skimmer's swing arm pivots from the rear of the chassis, with the polycarb tube + cup at the tip. The worst case is arm horizontal forward over a goal — the moment arm is at maximum and the entire payload mass is acting at the tip.

The math

Manipulator analysis — why polycarb tube wins for Skimmer

Unlike the Falcon and Heron analyses where pneumatic pincers won, on Skimmer the polycarb tube is the right answer. The reason is structural: Skimmer's swing arm has only one DOF (the arm rotation), and there's no wrist or chain bar providing independent end-effector orientation. The tube's rotation motor is the only orientation control on the manipulator end — not redundant with anything.

Pneumatic implementation — prototype progress + tank placement

The geometric insight — where the tanks can go

Skimmer’s arm sweeps a vertical plane through the chassis longitudinal centerline, with the arm tip tracing a 12″-radius arc above the chassis (pivot 12″ up, total arm + tube reach 12″). The lateral sides of the chassis are entirely clear of the arm’s path. That’s where the tanks belong.

One tank or two? — R25

R25a permits a maximum of two (2) air tanks. R25b caps pressure at 100 psi. For a single-cylinder cinch system actuating once per scoring cycle, one tank is probably enough — the team could save ~0.5 lb of mass and a chunk of chassis bay real estate.

Recommendation: bench-test with one tank first (less mass, simpler plumbing). Measure pressure at the end of a 2-minute scrimmage match. If the gauge reads below 60 psi at match-end, add the second tank. Document the decision and the measured pressures in EN4 — this is exactly the kind of empirical iteration the engineering notebook wants to see.

Pressure gauge placement — R26

The VEX pressure gauge is required and must be visible and readable to inspectors and referees without removing other Robot components or mechanisms. The tanks themselves can live inside the chassis bay, but the gauge needs to peek out somewhere visible. Easy solutions:

• Side-rail gap mount. A small bracket pokes the gauge through a gap in the side rail so the dial is visible from outside the chassis. Works well with side-rail-mounted tanks — short plumbing.
• Rear-face mount. Gauge mounted on the back face of the chassis, below the loader-catch tube position (so the tube doesn’t block it at Position B). Easy for inspectors to read while the robot sits on the table.
• Top of chassis, off-center. Mounted on the chassis top, away from the arm’s center-plane arc — specifically toward the side rails where the arm doesn’t reach. Most visible but most exposed to collision damage.

The first option (side-rail gap) is cleanest if the tanks are already on the side rails.

The cylinder placement question — arm vs chassis

This is the bigger design decision underneath the tank placement. The cylinder pulls the string that tensions the polycarb tube. Where the cylinder lives drives where the hoses and string have to route. Two options:

Hose routing for the on-arm cylinder option

If the team picks the on-arm cylinder, the pneumatic hose has to traverse the 230° arm arc without kinking or pinching. A few rules of thumb:

• Service loop length: hose should have at least 1.5× the linear pivot-to-cylinder distance in slack, so the loop expands/contracts through rotation without going taut.
• Loop position: the loop should hang off the side of the arm OPPOSITE the direction of sweep approach at Position B — otherwise the loop slaps against the chassis or loader during the catch.
• Cable carriers / tubing protection: R19e permits commercially-available bundling for pneumatic tubing protection — a cable track or spiral wrap is legal and recommended where the hose flexes hardest.
• Strain relief at endpoints: the cylinder fitting and the tank/solenoid fitting are the failure points if the hose tugs through rotation. A small bracket at each end that anchors the hose just past the fitting prevents the fitting from being stressed.

Plumbing layout (suggested for the on-arm cylinder option)

With both tanks tee’d together upstream of the regulator, the system shares pressure across both reservoirs. The gauge sits on the high-pressure side before the regulator if a regulator is used (matches R26a); otherwise it sits anywhere on the line after the tanks. Single solenoid with single cylinder is the simplest control architecture.

Open pneumatic questions for the team

1. Single-acting vs double-acting cylinder? Single-acting + spring return is simpler (only one hose), but the spring may resist tube release. Double-acting needs two hoses but gives crisp release.
2. Cylinder stroke length needed to fully cinch and release the tube around a cup? Bench-test with the prototype tube to measure.
3. String anchor point on the tube — epoxied loop? Through-hole with a knot? R27e prohibits gluing parts together, but a string with a knot through a drilled tube hole is fine.
4. Tank orientation — valve up or valve down? Valves up keeps debris out of the valve seat; valves down (lying horizontal under the chassis) saves vertical space but risks dirt ingestion.
5. How does the prototype 3D-printed tube translate to a cut polycarbonate version? Document the fabrication path in EN4: which curves, which cuts, how the bend is heat-formed (heating polycarb to aid bending is legal per R24d).

Game-ready analysis

How Skimmer performs in a 2-minute match: which goals it covers, how long each cycle takes, what the driver does, and the realistic cycle count per match. Skimmer’s defining advantage is the passive front-flex-wheel intake that runs while the robot is driving — this collapses the “drive to element + grip element” sequence into a single overlapping motion, which is why Skimmer ranks fastest on cycle time across the fleet.

Goal coverage

The polycarb tube is mounted along the swing arm. As the arm sweeps from Position A (-30° below horizontal) to Position B (+90° vertical), the tube center traces an arc 14″ from the pivot. With the pivot 12″ above the field tile, the tube can deposit at any point along the arc by stopping the sweep at the appropriate angle.

All three field goals are accessible mid-sweep. The team’s scoring routine doesn’t require lift extension — the natural sweep arc passes through every goal’s deposit height, so the driver doesn’t pause at intermediate positions; instead, they release the cinch as the tube swings through the right altitude.

Cycle time math

Two cycle types: floor pickup (drive over an element with the flex wheels running) and loader catch (back-end-first to the loader, vertical tube receives). Floor pickups are the fast cycle — this is where Skimmer outpaces every other architecture in the fleet.

The cycle-time delta against Pelican is decisive on floor pickups (~3 s vs ~6 s). On loader cycles the gap closes (~7 s vs ~10 s) because both architectures spend most of the cycle on field travel rather than mechanism actions. Skimmer’s strategy implication: prioritize floor cycles, use loader trips only for tall-goal cup deposits where the higher-value score justifies the cycle-time cost.

Auton routine (15 s)

Skimmer’s auton plays to its passive intake: drive forward, the flex wheels pick up alliance pins along the way, sweep to deposit. No precision claw alignment required.

1. 0.0–0.5 s: Pre-loaded element in tube at start. Arm starts at Position A (intake-ready).
2. 0.5–2.0 s: Drive forward over alliance pin row, flex-wheel intakes 1-2 pins.
3. 2.0–3.0 s: Sweep to alliance angle, release cinch — first deposit.
4. 3.0–4.5 s: Continue forward, flex-wheel intakes 2-3 more pins.
5. 4.5–5.5 s: Sweep, release — second deposit.
6. 5.5–8.0 s: Drive to a second pin row, intake while driving.
7. 8.0–9.5 s: Sweep, release — third deposit.
8. 9.5–15.0 s: Repeat. Position robot for driver-control handoff.

Realistic auton output: 4–6 alliance scores plus auton bonus. This is the ceiling-setter for Skimmer; 355Z’s passive intake reaches scoring rates that claw-based architectures cannot match in a 15-second window.

Driver-control routine (1:45)

The 105-second driver-control phase is where Skimmer compounds its cycle-time advantage. A representative driver pattern:

• 0:00–0:40 (sustained floor cycling, ~40 s): Drive across the alliance and short-goal zones, flex wheels pick up pins continuously. Sweep arm to deposit at whichever goal is closest. Target 10–13 floor cycles.
• 0:40–1:00 (loader trip, ~20 s): Drive back-end-first to alliance loader, arm at Position B vertical, human drops a cup, drive to tall goal, sweep + release. One high-value cup score.
• 1:00–1:30 (mixed cycling + toggle, ~30 s): Resume floor cycling. Activate field toggle when path crosses one (flex wheel takes 0.5–1 s of contact). Target 6–8 cycles plus 1–2 toggle activations.
• 1:30–1:45 (endgame, ~15 s): Retract arm to Position A, drive to endgame zone, hold position.

Realistic driver-control output: 17–22 cycles + 1–2 toggle activations. Combined with auton: 21–28 elements scored per match, plus toggle bonus.

Strengths

• Passive intake while driving. The single biggest cycle-time differentiator in the fleet. No stop-grip-resume sequence; intake overlaps with field travel.
• Single-DOF lift. One swing arm, one motor, one motion. Easier to tune than multi-link lifts; failure modes are concentrated and well-understood.
• Dual-feed through one manipulator. Floor pickups go through the flex-wheel-to-tube handoff at Position A; loader drops go through the tube’s open mouth at Position B. Same tube, same cinch, two ways to load it.
• Clear inspiration narrative. 355Z’s Simple Bot Override is documented and well-known; Skimmer’s adaptation makes for a strong notebook story (informed by prior art, then differentiated with documented reasoning).
• 355Z-validated geometry. The fundamental architecture is empirically proven by 355Z’s competition footage.

Limitations

• Polycarb tube fabrication risk. The tube is custom-fabricated (R24 plastic, 350° revolve, heat-bend procedure). Yield isn’t guaranteed first-try; the team needs a stock of practice blanks.
• Pneumatic cinch dependency. Tube cinch requires pneumatic control. Lose air mid-match, lose the manipulator. Backup: electric servo-driven cinch as fallback (more motor budget needed).
• Flex-wheel-to-tube handoff geometry. The Position A handoff has a ~1″ gap between flex wheels and tube; needs a back-stop or hood to ensure elements travel up into the mouth rather than getting kicked forward. Open question per build log.
• No multi-element capacity in tube. Tube holds one element at a time. Can’t pre-stage cup-on-pin combos for batch scoring.
• Single point of failure on the tube. If the tube cracks or a cinch line breaks, the entire scoring pathway is offline. Mitigation: bring a spare assembled tube to events.

Match scenarios

Decision matrix — Skimmer vs. fleet

How Skimmer compares to the other fleet bots, with each at its recommended manipulator (Skimmer with tube, Falcon with pincers, Heron with pincers, Osprey with claw):

CAD starting point

Subassembly dimensions

Reach math — can the swing arm cover all three goals?

Universal manipulator mounting plate — the swing arm tip interface

Skimmer's swing arm tip uses the same universal mounting plate spec as the rest of the fleet (Heron, Crane, Stork, Falcon). This means the polycarb tube manipulator the team is building for Skimmer can drop directly onto any other architecture in the fleet without re-CADing — and vice versa, if the team decides to test a pneumatic pincer or V5 claw, those drop-in adapters are already specified.

Open questions for the team

• Tube mounting axis. Tube axis along the arm length (mouth direction rotates with arm) or perpendicular (mouth direction stays constant)? The dual-feed concept assumes axis-along-arm. Verify on the prototype.
• Flex-wheel array height + count. Wheel-bottom height assumption: touching the tile or ~0.25″ above. Pins lying flat are ~1.6″ thick; the wheels need to compress on the top of the pin without the chassis lifting off the tile under reaction force. Wheel count: 7 is the working assumption matching the 355Z reference; 6 is acceptable if axle space tight, 8 if more grip area is needed. Verify both height and count from physical mock-up before locking the axle bracket geometry.
• Front intake motor power: 5.5W half-motor or 11W full motor? Flex wheels driving against the floor have higher torque demand than free-spinning flex rollers. 5.5W preserves a 5.5W spare in the motor budget but may be marginal at low RPM under load; 11W maxes out the 88W cap (R10a) but provides headroom. Bench-test both. Drive the wheel array against a static cup with the robot weight on it; measure stall current. If 5.5W stalls or thermal-trips, use 11W and accept zero spare budget. Note: chain reduction from motor to axle (e.g., 1:2) shifts the torque-vs-speed tradeoff and may make 5.5W viable.
• Flex-wheel-to-tube handoff geometry. With the swing arm at Position A, the tube hovers directly above and slightly behind the flex-wheel array. The wheels’ rotation grips elements via friction with the tile and rolls them backward into the chassis, where they need to be redirected upward into the tube’s open mouth. Open question: does the intake need a back-stop, hood, or shaped funnel above the wheel array to redirect rolled-in elements up into the tube rather than letting them slide along the chassis floor or jam on the tube’s lip? 355Z’s design likely solves this — verify by watching the reference video below and looking specifically at the geometry above the wheel row.
• Swing arm rubber-band verification. The torque section above sizes the assist at 6 × #64 bands (~7.5 lbf at 5″ perpendicular) for a τ_demand of 19 lb-in. Verify on the prototype: does the arm hold horizontal under cup+pin payload without motor power? Does it sag at the rest position? Does the driver report the arm "popping up" off the loader? Iterate band count by ±2 based on actual feel.
• Flex-wheel toggle interaction. Push the toggle (spinning flex wheel deflects it) or grip-and-rotate (flex wheel grabs and rotates the toggle)? Push is simpler; grip is more controllable. Test both.
• 355Z reference. Watch the 355Z Simple Bot Override video again with the build team. Note any details Skimmer needs to deviate from (e.g., 355Z's claw → Skimmer's tube). Document deviations in the engineering notebook with reasoning.

Port map (template — fill in as built)

Build log

Each build session adds an entry: date, team members, what was attempted, what worked, what didn't, decisions made.

Engineering notebook references

Cross-references between this page and the team's physical EN4 notebook. As each notebook entry is written, fill in the page number to maintain bidirectional traceability.

Inspiration · 355Z Simple Bot Override

Skimmer’s architecture is heavily inspired by Team 355Z’s “Simple Bot Override · 2 Bar” design — a fast-scoring two-bar swing arm with a flex-wheel intake, optimized for cycling pins and cups quickly through low-complexity mechanics. The geometry choices on this page (tower-mounted swing arm, two rest positions, front-flex-wheel intake feeding into the polycarb tube) trace directly to 355Z’s build.

📺 Watch on YouTube → · 7:47 runtime · Team 355Z

Watch with the team specifically for:

• The two rest positions and how the driver transitions between them. Verify Skimmer’s interpretation (Position A forward-down at -30° for flex-wheel handoff, Position B vertical at +90° for loader catch) matches 355Z’s actual usage. Note the angles 355Z chose — we’re assuming -30° and +90°, but 355Z may have settled on different values empirically. In particular, watch whether 355Z uses arm-vertical for the loader catch or some other angle.
• How the front flex-wheel array hands elements off to the polycarb tube. Look for a back-stop, hood, deflector, or shaped guide above the wheel row that funnels elements upward into the tube. This is Skimmer’s “biggest unsolved subsystem” per the open questions above — 355Z likely has the answer.
• Flex-wheel count and spacing. The reference image shows a row of wheels on a single axle. Count them carefully — team assumption is 7, but 355Z may use 6 or 8 depending on chassis-front geometry and axle stub length. Note also the wheel diameter (likely 2″), durometer (the team is using 40), and any spacers between wheels.
• Sweep speed and PID tuning behavior. Watch for overshoot at deposit positions, controlled deceleration approaching each rest position, and the timing between intake and score. Skimmer’s 11W Red 100 RPM swing motor with 1:5 reduction has a specific top speed; estimate 355Z’s and compare.
• Tube mounting axis (along-arm vs cross-axis). See how the tube’s spatial orientation changes through the sweep. The orientation question on this page assumes axis-along-arm, but 355Z’s design may differ.
• Toggle activation pattern. When does 355Z engage the toggle relative to scoring cycles? Is it a separate dedicated motion, or does the swing arm sweep past the toggle on the way to a goal?
• Cycle time per scoring action. Time how long 355Z takes from intake to deposit. That number is Skimmer’s benchmark. If Skimmer’s prototype is significantly slower, identify which subsystem is the bottleneck and iterate.

Open architectural question — dual-side toggle?

Skimmer currently specs a single 5.5 W flex wheel toggle. The new Osprey (2822E) dual-side architecture (2 × 5.5 W mirrored, independent L/R driver controls) raises the question whether Skimmer should reconsider before Phase A scrimmages.

See also

• Polycarb tube design study — the manipulator's full design rationale; Skimmer is the architectural home this study was waiting for.
• Polycarb tube CAD guide — OnShape modeling of the tube subassembly.
• Polycarb tube fabrication guide — bench reference for the heat-bend procedure.
• Pelican — Team 2822C's simultaneous build (Hero Bot baseline four-bar architecture). The offensive-pace vs cycle-reliability counterpart to Skimmer.
• Spoonbill — Team 2822D’s simultaneous build (Hero Bot four-bar with center tower + rotating V5 claw). The orientation-precision counterpart in the Hero Bot-derived family. Different lineage from Skimmer (which is 355Z-derived, not Hero Bot-derived).
• Osprey — Team 2822E's simultaneous build (Hero Bot-derived chain-bar variant). Loader-only committed cycle; tall-goal optimized. The other architectural counterpart to Skimmer's floor-pickup pace.
• Kite — Team 2822F's simultaneous build (Hero Bot-derived four-bar variant with shared-shaft power transmission). Different lineage from Skimmer (Hero Bot baseline, not 355Z) but interesting for comparison: where Skimmer extracts speed from a single-DOF swing arm + passive intake, Kite extracts wattage headroom from sharing motors across mechanisms. Both are the “clever architecture” bots of the fleet in different ways.
• Owl — Team 2822O's simultaneous build (Hero Bot-derived four-bar with a pneumatic-pivot flex-wheel intake on the Spoonbill chassis). Different lineage from Skimmer but the only other bot in the fleet running an active rolling intake as a manipulator. Useful comparison for flex-wheel intake mechanics: Skimmer’s row of intake wheels sit at the chassis front in a fixed orientation; Owl’s pair of wheel pairs sit on a pivoting body at the arm tip. Same primitive (active flex-wheel rolling intake), different application.
• Spartan Hero Bot V1.5 — the four-bar baseline that Pelican and Osprey both descend from. Different lineage from Skimmer (which is 355Z-derived, not Hero Bot-derived).
• Falcon — the 4-DOF arm concept (formerly Mantis). More articulated but more complex; Skimmer's foil.
• Heron — stacked four-bar + chain bar concept. More vertical reach; Skimmer is faster.
• Crane — six-bar vertical lift concept.
• Stork — reverse four-bar (DR4B) concept.