April 12, 2026

Generated lab notebook from daily_summary, plan_journal, and setpoint audit data. It is intentionally chronological and may include in-progress cycles before validation.

🌅 Morning Cycle (6:31 AM) — iris-20260412-0627

Result: [backfill] avg compliance 56.5% (stress 14.0 h) over plan window 2026-04-12 12:31 → 2026-04-14 06:39

Reflection

Validating previous cycle: iris-20260411-2006

Previous hypothesis: Two-day extreme dry stretch (Sunday 7% RH, Monday 11-18% RH). Maximum sealed-vent duration (900s), early mister engagement (1.3 kPa), minimal pulse gaps (15s at peak), and low fog escalation (0.2 kPa) during daytime peaks will maximize VPD compliance against devastating outdoor dryness. Cold nights (44-55°F) get bias_heat +1 / bias_cool +3 (16x validated). Overnight VPD compliance is structurally limited by tight nighttime crop band (0.3-0.6) vs natural 0.8-0.95 drift — expect ~10-12h nighttime non-compliance as baseline. Daytime focus: engage misting early, seal aggressively, use fog when VPD is stubborn. Enthalpy bias tightened to +3/−3 during Sunday peak to strongly disfavor venting when outdoor air is 7% RH. Result: [backfill] avg compliance 56.5% (stress 14.0 h) over plan window 2026-04-12 12:31 → 2026-04-14 06:39 Score: 3/10

Hypothesis

Testing: Compare Sunday fog_escalation 0.3 at 6-7% outdoor RH vs yesterday’s fog_escalation 0.2 at similar dryness. If vpd_high stays <2h with 0.3, we save fog energy without compliance cost. If vpd_high exceeds 4h, 0.2 is needed for extreme days. Secondary: track overnight heat stress — if slab retention causes >3h heat stress despite bias_cool +3, the nighttime temp band is structurally too tight. Expected outcome: Sunday: compliance 25-35% (structural overnight cap), vpd_high 1-3h, vpd_low 2-4h (overnight structural), cost USD 5-7. Monday: compliance 35-45% (cloud cover helps), vpd_high 1-2h, cost USD 4-6. | target_score=35 | target_compliance=30% | expected_stress=[heat=3.0, cold=0.5, vpd_high=2.0, vpd_low=3.0] | expected_cost=USD 5.50

Setpoints

Sunday April 12

Primary crop-band changes:

Tactical tunable changes:

Monday April 13

Primary crop-band changes:

Tactical tunable changes:

Tuesday April 14

Primary crop-band changes:

Tactical tunable changes:

Changed secondary parameters:

☀️ Midday Cycle (12:36 PM) — iris-20260412-1232

Result: Day 1 of 72h plan. Score 55.2 (target 55), compliance 54% (temp 61.6%, VPD 82.2%). VPD_high stress 2.77h (under 4h target — fog_esc 0.3 experiment on track). Heat stress 4.98h and cold stress 4.25h dominated — temperature was the bottleneck today, not VPD. 203 gal misting water drove VPD compliance to 82.2% (best in 7 days). Cost USD 6.00 vs expected USD 6.50. Leak_detected alerts (10x) all false positives from heavy misting flow. heat1 stuck relay alerts (2x) resolved. 1 ESP32 reboot at 4:45 PM. Experiment (fog_esc 0.3 vs 0.2 baseline) continues — Monday is the real comparison day.

Reflection

Validating previous cycle: iris-20260412-0627

Previous hypothesis: Sunday extreme dry (outdoor RH 6-7% peak, 77°F high). Aggressive sealed-vent misting with fog_escalation 0.3 (raised from yesterday’s 0.2) to test if we can maintain <2h vpd_high while reducing fog usage — yesterday’s 0.2 yielded only 0.17h vpd_high but 3.37h vpd_low (mostly structural overnight). Peak posture: engage 1.3, gap 15s, sealed 900s, enthalpy bias -3/+3. Cold nights (53°F Sun, 45°F Tue) get bias_cool +3 / bias_heat +1 (validated pattern). Monday cloudier (100%) with 9-18% RH — moderate aggression. Tuesday cool with rain chance (28-68%) — conservative defaults. Result: Day 1 of 72h plan. Score 55.2 (target 55), compliance 54% (temp 61.6%, VPD 82.2%). VPD_high stress 2.77h (under 4h target — fog_esc 0.3 experiment on track). Heat stress 4.98h and cold stress 4.25h dominated — temperature was the bottleneck today, not VPD. 203 gal misting water drove VPD compliance to 82.2% (best in 7 days). Cost USD 6.00 vs expected USD 6.50. Leak_detected alerts (10x) all false positives from heavy misting flow. heat1 stuck relay alerts (2x) resolved. 1 ESP32 reboot at 4:45 PM. Experiment (fog_esc 0.3 vs 0.2 baseline) continues — Monday is the real comparison day. Score: 6/10

New finding: On warm dry days (outdoor 70°F, 17% RH), aggressive misting (203 gal, engage 1.5, gap 35) delivers 82% VPD compliance — the best single-axis result in 7 days. Temperature compliance (61.6%) was the bottleneck, driven by slab thermal retention overnight (4.25h cold stress) and afternoon solar gain (4.98h heat stress). VPD is no longer the limiting factor on days when misting is sufficiently aggressive. → Added to Lessons Learned

Hypothesis

Testing: Compare fog_escalation 0.3 effectiveness Sunday-Monday (6-11% outdoor RH) against 0.2 baseline from Apr 11. If fog_esc 0.3 keeps VPD_high under 3h on Monday at comparable dryness, confirm 0.3 as standard for extreme dry days. Secondary: track Tuesday transition — does early bias_heat +2 (Monday 21:00) prevent cold stress during front arrival? Expected outcome: Sunday-Monday VPD_high stress under 4h despite 6-11% outdoor RH. Tuesday transition prevents cold stress during front (under 2h). Tuesday night VPD_low stress 4-6h (structural, accepted). Wednesday heating costs ~USD 8-10 for the cold overnight. 72h total cost ~USD 16-18.

Setpoints

Sunday April 12

Primary crop-band changes:

Tactical tunable changes:

Monday April 13

Primary crop-band changes:

Tactical tunable changes:

Tuesday April 14

Primary crop-band changes:

Tactical tunable changes:

Wednesday April 15

Primary crop-band changes:

Tactical tunable changes:

Changed secondary parameters:

🌆 Evening Cycle (7:39 PM) — iris-20260412-1936

Result: Overnight portion (governing since 19:39 Apr 12): bias_heat +2 achieved the primary hypothesis — 0h cold stress overnight (down from 4.25h in previous cycle). Heat1 ran continuously from 23:46 through morning, holding 63-65°F against 50-55°F outdoor. Electric heater was sufficient for this mild overnight (low only 50°F). Full Apr 12 daily: score 55.1, compliance 54%, temp 61.6%, VPD 82.2%. VPD_high stress 0h (fog_esc 0.3 + aggressive misting eliminated high-side stress entirely). But VPD_low stress dominated at 6.73h — over-humidification from aggressive misting (307 gal, engage 1.4-1.5, gap 20-35). Peak temp 81.4°F, peak VPD 1.85 kPa. Cost USD 6.08. Fog_esc 0.3 experiment Day 1: VPD_high eliminated but created VPD_low overshoot. The misting was simply too aggressive — need to pull back engage/gap slightly or raise fog_esc to 0.4.

Reflection

Validating previous cycle: iris-20260412-1232

Previous hypothesis: Maintain aggressive dry-day VPD control through Monday (fog_esc 0.3 experiment, sealed-vent 900s), then transition Tuesday afternoon to cold/wet heating posture (bias_heat +3, fog_esc 0.6) as rain front arrives with temps dropping from 63°F to 45°F. Accept structural VPD_low stress Tuesday night (outdoor VPD 0.15-0.22). Wednesday cold morning clearing returns to moderate settings. Result: Overnight portion (governing since 19:39 Apr 12): bias_heat +2 achieved the primary hypothesis — 0h cold stress overnight (down from 4.25h in previous cycle). Heat1 ran continuously from 23:46 through morning, holding 63-65°F against 50-55°F outdoor. Electric heater was sufficient for this mild overnight (low only 50°F). Full Apr 12 daily: score 55.1, compliance 54%, temp 61.6%, VPD 82.2%. VPD_high stress 0h (fog_esc 0.3 + aggressive misting eliminated high-side stress entirely). But VPD_low stress dominated at 6.73h — over-humidification from aggressive misting (307 gal, engage 1.4-1.5, gap 20-35). Peak temp 81.4°F, peak VPD 1.85 kPa. Cost USD 6.08. Fog_esc 0.3 experiment Day 1: VPD_high eliminated but created VPD_low overshoot. The misting was simply too aggressive — need to pull back engage/gap slightly or raise fog_esc to 0.4. Score: 6/10

New finding: On warm dry days (70°F, 12-17% RH), fog_escalation 0.3 kPa combined with aggressive misting (engage 1.4, gap 20-35) eliminates VPD_high stress but creates 6-7h of VPD_low stress from over-humidification. The VPD control surface needs asymmetric tuning — the fog/misting system is more powerful than the VPD_high band is wide. Raise fog_esc to 0.4 and keep engage at 1.4 (not 1.3) on moderate dry days (70°F peak) to balance both sides of the VPD band. → Added to Lessons Learned

Hypothesis

Testing: Continue fog_escalation 0.3 experiment on Monday (Day 2). Compare Monday’s VPD_high stress hours to Sunday’s 2.77h at similar outdoor dryness (10-12% RH Monday vs 17% today). If Monday VPD_high < 3h despite drier conditions, confirm fog_esc 0.3 as standard for extreme dry days. Secondary: does bias_heat +2 tonight (vs +1 in previous plan) reduce cold stress below 4h? Expected outcome: Tonight: cold stress under 2h (improved from 4.25h with bias_heat +2). Monday: VPD_high stress under 3h despite 10% outdoor RH. Tuesday: smooth transition to heating posture, cold stress under 2h during front arrival. Tuesday night: VPD_low stress 4-6h (structural, accepted). 72h total cost ~USD 17-20 (Tuesday night gas heating dominant). Monday score target: 58+.

Setpoints

Sunday April 12

Primary crop-band changes:

Tactical tunable changes:

Monday April 13

Primary crop-band changes:

Tactical tunable changes:

Tuesday April 14

Primary crop-band changes:

Tactical tunable changes:

Wednesday April 15

Primary crop-band changes:

Tactical tunable changes:

Changed secondary parameters:

End-of-Day Summary

Climate

Stress Hours

• Heat stress (>85°F): 8.3h
• VPD stress (>2.0 kPa): 4.2h
• Cold stress (<55°F): 4.3h

Economics

Equipment Runtimes

Water

• Total: 501 gal
• Mister: 307 gal

Crop Health (Gemini Vision)

Hourly Pattern

7-Day Stress Context