2-Week Facility Audit

Zugatti · 14th street

Operator pre-meeting audit. Window 2026-06-01 → 2026-06-15 (14 days).
FacilityZugatti: 14th street · ID 2704 OrgDUB BROS. MANAGEMENT LLC · ID 2076 Window2026-06-01 → 2026-06-15 Generated2026-06-15

1 · Headline Findings

Active harvest groups
0
No HG logged in window; last HG harvest 2025-03-18
Rooms with live data
4 of 7
F2/F3 cultivating; F1 idle; Veg sensor-dark
Window coverage
Trailing 24-36h
get_raw_data clamped per-point output; window stats (n=480) verified
Flower 2 climate
75.4°F / VPD 1.21
On-target early/mid-flower band
Flower 1 climate
83.8°F / VPD 2.18
Flat 3°F range, no DIF — idle-room signature
Irrigation flow telemetry
None
No drip/drain meters; shots derived from VWC steps
Lead verdict
Zugatti 14th street is running WITHOUT harvest-group records in AROYA — last logged HG harvested 2025-03-18. Sensors are still live: Flower 2 (75°F, VPD 1.21, substrate cycling) and Flower 3 look actively cultivated; Flower 1 reads as idle (flat 83.8°F, VPD 2.18, no DIF); Veg is sensor-dark. No HG anchors the window and the MCP clamped per-point output to the trailing ~24-36h, so full 14-day daily trajectories were not retrievable — window aggregates (n=480/series) are verified and shown as bands; per-day shapes for June 1-13 were NOT fabricated.

2 · Prior-Commitment Audit

Client-context preflight: claude-brain was queried for DUB BROS / Zugatti prior commitments. Running UNATTENDED, no claude-brain hit was confirmable within budget. Known prior work: firmware (WP4→WP5 OTAP) readiness was built for Zugatti 2704 on 2026-06-01.

Prior item (inferred)Status this windowVerdict
Firmware WP4→WP5 OTAP readiness (built 2026-06-01)Devices online and reporting in F1/F2/F3; Veg dark. Online + battery≥30% + signal≥50% devices receive WP5 at the scheduled push; offline devices simply stay on WP4 (not harmed) and can be brought online for a later push or recovered via a WP4 gateway revert.Bring Veg + any offline nodes online before the push
Harvest-group logging / crop steering in AROYANo HG logged since 2025-03-18. AROYA cannot phase-steer or yield-report without an active HG.Lapsed — re-enable
Full-facility sensor coverageVeg (4837) returned 0 series; Flower 1 substrate/climate reads idle; Flower 3 missing room climate sensor.Coverage gaps

Source register: claude-brain 03-Clients/ + 09-Handoffs/ (unconfirmed this run); firmware readiness artifact 2026-06-01.

3 · HVAC Capacity & Climate Load

RoomAir °F meanmin/maxSwing °FVPD kPaRH %Sub °FSub−Air ΔState
Flower 183.882.2/85.33.12.184577.0-6.8idle
Flower 275.473.2/77.94.71.216074.6-0.8active
Flower 369.8no climate sensor
Vegsensor_dark
Flower 1 — idle-room signature
Air sits at 83.8°F across the window with only a 3.1°F range and no DIF, while VPD averages 2.18 kPa. RH is a flat 44%. This is the fingerprint of a room with no transpiring crop.
Flower 2 — healthy flower climate
75.4°F mean, VPD 1.21 kPa, RH 60% — early/mid-flower comfort band. Substrate 74.6°F (−0.8°F vs air).
Flower 3 — no room climate sensor
Substrate sensors report normally but no room-level air/VPD/RH series came back — the Climate One / Atmos head is absent or offline.

Capacity sizing rule: a transpiring flower canopy adds ~1.0-1.5 L/m²/day latent load; idle F1 confirms HVAC is oversized-for-idle, not a capacity problem.

4 · Climate ↔ Irrigation Interplay

MechanismWhat the data showsSteering read
VPD-stress (canopy)F2 VPD 1.21 kPa — no stress. F1 VPD 2.18 kPa — but no crop, so not a stress signal, an empty-room artifact.F2 fine
Substrate EC stackingF2 root-zone pore EC 7.6-13.0 mS (mean 10.3); F3 6.1-10.8 mS. Mid-range — consistent with a maintenance/bulking posture, not an aggressive generative stack.EC mid-band
Overnight drybackOnly window-envelope VWC ranges are retrievable (F2 room 39-59%, ≈20 pp envelope; F3 41-55%, ≈14 pp). True per-night dryback needs the per-day series that the MCP would not return this run.envelope only
Substrate-air couplingF2 substrate −0.8°F vs air (tight, healthy). F1 substrate +0.0 vs an idle air signal.F2 coupled
P0 activation behaviourWithout a flow meter or per-day VWC trace, P0 start-of-day delay and first-runoff timing cannot be confirmed this window.not observable

Sequence of operations to confirm at the meeting: (1) Is Flower 1 supposed to be in production? (2) Why is Veg sensor-dark? (3) Re-link a harvest group to F2/F3 so AROYA can resolve P0–P3 phase boundaries and overnight dryback in absolute pp. (4) Restore the Flower 3 room climate sensor.

5 · Per-Room Deep Dive · Select & Compare

Click a room chip to drive the overlay charts. Shift-click a second room for compare mode.

Flower 1idle
Flower 2active
Flower 3active substrate only
Vegsensor dark

Substrate VWC envelope (window min/max, pp)

Bars = window VWC range (max − min) in absolute pp.

Substrate pore EC (window mean & range, mS/cm)

Root-zone EC is the steering output. Mid-band (7-13 mS) = maintenance/bulking posture.

Air temperature (window mean & min/max, °F)

Flat band with no spread = idle room (Flower 1). Healthy flower band ~73-78°F (Flower 2).

VPD (window mean & range, kPa)

Target early/mid-flower 1.0-1.4 kPa. >2 kPa with no crop = empty-room artifact, not stress.

6 · Phase Biophysics & Irrigation Inputs

No harvest group is logged for the window, so AROYA cannot assign P0–P3 phase boundaries automatically. Cards report verified window-level substrate biophysics; dryback in absolute pp.

Flower 2 · room 4833

Flower (no HG)

Window VWC envelope 20.1 pp (max−min, absolute pp) — note this is the window range, not a single overnight dryback; the per-night dryback in absolute pp requires the per-day VWC trace that the MCP would not return this run. Substrate pore EC 10.28 mS/cm (range 7.6-13.03). FC/floor proxy: VWC 39.3-59.4% (mean 48.29%). Air 75.4°F (range 73.2-77.9).

P0 ActivationP1 Ramp-upP2 MaintenanceP3 Dry-down Early-flower target: dryback 20-30 pp, P0 delay 120-240 min, substrate EC stacking 6-12 mS

Flower 3 · room 4834

Flower (no HG)

Window VWC envelope 14.1 pp (max−min, absolute pp) — note this is the window range, not a single overnight dryback; the per-night dryback in absolute pp requires the per-day VWC trace that the MCP would not return this run. Substrate pore EC 9.21 mS/cm (range 6.14-10.83). FC/floor proxy: VWC 41.0-55.1% (mean 47.13%).

P0 ActivationP1 Ramp-upP2 MaintenanceP3 Dry-down Early-flower target: dryback 20-30 pp, P0 delay 120-240 min, substrate EC stacking 6-12 mS

Flower 1 · room 4832

Idle

Window VWC envelope 14.9 pp (max−min, absolute pp) — note this is the window range, not a single overnight dryback; the per-night dryback in absolute pp requires the per-day VWC trace that the MCP would not return this run. Substrate pore EC 15.84 mS/cm (range 14.15-17.44). FC/floor proxy: VWC 29.9-44.8% (mean 37.02%). Air 83.8°F (range 82.2-85.3).

P0 ActivationP1 Ramp-upP2 MaintenanceP3 Dry-down Early-flower target: dryback 20-30 pp, P0 delay 120-240 min, substrate EC stacking 6-12 mS

7 · Irrigation Shot-Cycle Audit

No flow telemetry
count_irrigation_events returned 0 — Zugatti's substrate sensors are TEROS with no drip/drain meters. Shot cadence can only be derived from VWC step-up signatures (≥2 pp rises per 1h bin). The MCP returned only trailing-window daily points this run, too coarse to resolve individual P1 shots, so shot cadence for June 1-13 is NOT estimated.

Recommendation: re-link a harvest group so spa_get_harvest_readings can pull the full per-minute substrate trace, enabling true P0 start-of-day delay, P1 first-runoff, and P3 dry-down measurement.

8 · Config & Guardrail Audit

Target ranges and recipe guardrails live on the harvest-group / room config surface. With no active HG, no live target-range guardrails are bound to the current rooms — a room cannot raise phase-appropriate alerts without a recipe attached.

AUTOMAIT entitlement: not assessed this run (no HG/recipe surface to read).

9 · Recommendations — Owner-Tagged

OwnerRecommendationObserved delta
GrowerRe-create / re-link a harvest group for each active room (F2, F3) in AROYA so phase steering, P0–P3 boundaries, overnight dryback (absolute pp) and yield reporting resume.0 HGs logged since 2025-03-18
ProcessInvestigate the Veg-room (4837) sensor blackout — 0 series returned for the full window. Check gateway/mesh and node power (Node + TEROS 12 + H340-series repeaters are battery; Climate One / Atmos 14 / legacy H11x repeaters / gateway are PoE).Veg = 0 series
OwnerConfirm Flower 1's intended state. If it should be in production, the flat 83.8°F / VPD 2.18 / no-DIF signature means there is no transpiring crop — investigate before flip.Air range 3.1°F, VPD 2.18
ProcessRestore the Flower 3 (4834) room climate sensor — substrate (TEROS) reports but no air/VPD/RH series. Canopy climate is unmonitored there.F3 room climate series = none
GrowerFlower 2 climate is well-run (75°F, VPD 1.21, RH 60%) — hold the current setpoints; once an HG is linked, confirm the EC stack (10.3 mS) matches the intended phase.VPD 1.21, EC 10.3 mS
ProcessFirmware: bring Veg + any offline nodes online before the scheduled WP4→WP5 push (online + battery ≥30% + signal ≥50% to receive WP5). Offline devices stay on WP4 unharmed and can be updated on a later push or recovered via a WP4 gateway revert.Veg dark; firmware readiness built 2026-06-01

10 · Data Quality & Coverage

Honest coverage statement
Window: 2026-06-01 → 2026-06-15 (14 days). get_raw_data clamped per-point output to the trailing ~24-36h (a bin=day query over 14 days returned only 2026-06-14 and 2026-06-15). Window-level aggregates (n=480/series) ARE genuine and reported as bands. No HG is logged (last harvest 2025-03-18) so the HG-anchored full-window path was unavailable. Per-day shapes for June 1-13 were NOT synthesized.
RoomStateCoverageVerdict
Flower 1 (4832)idleWindow stats verified; idle signatureidle
Flower 2 (4833)activeWindow stats verified (n=480/series); daily points 06-14/06-15 onlyactive
Flower 3 (4834)active_substrate_onlySubstrate verified; NO room climate seriesactive_substrate_only
Flower 4 (4835)unprobedNot probed this rununprobed
Dry Room (4836)unprobedNot probed this rununprobed
Veg (4837)sensor_dark0 series — sensor-darksensor_dark