Second cultivation audit for HK Cassandra — graded against the 2026-06-24/25 baseline and AUTOMAIT install. The dominant signal this window is a facility-wide deep overnight dryback running 38–46 absolute pp in nearly every room regardless of phase, plus a correction to the June-24 "coco leaching" finding. Companion mesh report: Wirepas Mesh Health (2026-06-24).
| Commitment / finding (6/24–25) | Status this cycle | Evidence |
|---|---|---|
| R2 & R5 "coco leaching" (drip≫drain) | Corrected / resolved | Now reads drip<drain everywhere (R2 2.34/4.18, R5 2.14/4.47) = normal stacking. Earlier reading = swapped sensor labels (fix logged 6/25). |
| R5 channeling (~53 pp, 0.5 GPH ×2 emitters) | Still deep | R5 dryback still 46 abs pp (down slightly from ~53). Confirm emitter count / shot volume on Zone 1 (manual) and Zone 2 (AUTOMAIT). |
| R5 AUTOMAIT install (Zone 2, VEG 10) | Live & executing | Running since 6/25; room reporting 98% coverage. Room-avg dryback still deep — review Zone-1 manual schedule. |
| R10 feed EC ~1.54 mS (veg-strength) | Worsened | Feed EC now ~1.0 mS; substrate/drain stacked to ~11 mS. Escalated to a headline watch. |
| Cohort | Rooms | Plants | Phase (day) | Harvest | Cultivars |
|---|---|---|---|---|---|
| May-10 Late Flower | R3 · R9 · R10 · R11 | 1,425 | Late Flower (day 3 of phase) | 2026-07-13 (~12d) | Blue Nerdz · G-Runtz · MOB |
| Jun-3/4 Mid Flower | R1 · R2 · R5 | 1,728 | Mid Flower (day 6–7) | 2026-08-05/06 (~35d) | GX · Blue Nerdz |
| Jun-28 Early flower | R4 · R6 · R8 | 1,455 | Early Day 1-21 (day 3) | 2026-08-30 (~60d) | Blue Nerdz · BG×BN · BN×Mochi |
| Metric | Early flower / stretch (Jun-28) | Mid / bulking (Jun-3) | Late / ripening (May-10) |
|---|---|---|---|
| Overnight dryback (absolute pp) | 6–10 pp | 12–18 pp | 25–35 pp |
| Substrate pore EC | 3.0–5.0 mS (building) | 4.0–7.0 mS | 6.0–8.0 mS restack |
| Feed (drip) EC | 2.0–3.0 mS | 2.5–3.5 mS | 3.0–4.0 mS |
| Lights-on VPD | 1.0–1.2 kPa | 1.2–1.4 kPa | 1.4–1.6 kPa |
| Lights-on air temp | ~78–82°F | ~76–80°F | ~72–76°F (cool finish) |
P0/P1/P2/P3 = the 4-phase irrigation day the controller is programmed in: P0 activation (lights-on → first shot), P1 ramp-up to field capacity + first runoff, P2 maintenance pulses holding VWC, P3 dry-down (last shot → next lights-on) — the primary steering lever. Deep overnight dryback = a long, aggressive P3.
The three cohorts run visibly different climates, which is directionally correct: the May-10 ripening cohort runs cool (R3 air 66°F / soil 64°F, R10 70°F) to drive finish; the Jun-3 bulking cohort runs warm (R1/R2 ~82°F lights-on); the Jun-28 early cohort runs warmest and most humid (R8 84°F / RH 74%). The mismatch is that the deep P3 dryback is applied uniformly on top of these climates — a 44–46 pp dryback in a warm bulking room (R1/R2/R5) with lights-on VPD only ~1.0–1.2 kPa means the plant is being asked to pull a lot of water back from a substrate while transpirational demand is moderate, which lengthens the time at low VWC. R2 in particular pairs 45 pp dryback with a soft 1.03 kPa VPD and ~70% RH — raise VPD toward 1.2–1.4 or shorten P3 to keep the bulking rooms turgid.
Click a room chip for its panel. Click Compare to overlay rooms on shared charts. Dryback is absolute percentage points (peak − trough VWC). Colors follow the phase palette (rust = ripening, violet/blue = bulking, green = early flower).
feed(drip) EC 1.24 mS · substrate/drain EC 7.43 mS · drip<drain = normal stacking
substrate pore EC 5.42 mS (drip/drain split not separately pulled this run)
feed(drip) EC 1.0 mS · substrate/drain EC 10.97 mS · drip<drain = normal stacking
substrate pore EC 4.95 mS (drip/drain split not separately pulled this run)
feed(drip) EC 1.9 mS · substrate/drain EC 4.99 mS · drip<drain = normal stacking
feed(drip) EC 2.34 mS · substrate/drain EC 4.18 mS · drip<drain = normal stacking
feed(drip) EC 2.14 mS · substrate/drain EC 4.47 mS · drip<drain = normal stacking
substrate pore EC 2.59 mS (drip/drain split not separately pulled this run)
substrate pore EC 4.36 mS (drip/drain split not separately pulled this run)
feed(drip) EC 1.25 mS · substrate/drain EC 1.64 mS · drip<drain = normal stacking
Across all cohorts the substrate is being taken through a large daily VWC swing (peak ~70–79%, trough ~14–34%). In the ripening cohort this concentrates the pore EC (R10 to ~11 mS, R3 to ~7.4 mS drain) and drives finish. In the bulking cohort the same swing works against turgor and cell expansion — bulking wants VWC held near field capacity with small, frequent P2 pulses. In the fresh Jun-28 cohort, drying a barely-rooted block to a 14–24% trough (R8/R6) risks stalling the stretch. The steering intent that the numbers imply is generative everywhere; that intent only matches the phase for the May-10 rooms.
Leading with overnight dryback (the most diagnostic steering number): the facility is running a long, aggressive P3 dry-down in every room. Feed (drip) EC is on the low side (1.0–2.3 mS) so the substrate stack is being built by concentration under that dry-down rather than by feed strength — visible as drain EC sitting well above drip EC in every room (normal direction, but the spread is widening fastest in R10 and R3). Actions: (1) on the Jun-3 bulking rooms, add a late-day P2 maintenance shot to lift the overnight trough and pull dryback from ~45 pp toward 15 pp; (2) on the Jun-28 early rooms, shorten P3 and raise feed EC into the 2.0–3.0 mS band so the young stack builds on feed, not stress; (3) on R10, decide feed-up vs. flush given ~11 mS substrate and 12 days to harvest.
/dataReadings/ (ROOM grouping) at native ~3-minute cadence, window 2026-06-17 → 2026-07-01. Every flower room returned ≥98% timestamp coverage of the 14-day window (R6/R8/R4 span from their 2026-06-28 flip). Daily aggregates computed locally; lights-on/off split by the 12 highest-air-temp hours per room.