September 12, 2026

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

Planner Execution Ledger

Planner checkpoints can acknowledge that the active plan is still suitable without writing a new public plan ID. Those no-change decisions are part of the audit trail.

🌅 Morning Cycle (12:18 AM) — iris-20260912-0017

Reflection

Validating previous cycle: iris-20260911-0639

Previous hypothesis: {“conditions”:{“outdoor_temp_peak_f”:92.0,“outdoor_rh_min_pct”:8.0,“solar_peak_w_m2”:704.0,“cloud_cover_avg_pct”:50.0,“notes”:“Forecast-corrected hot-dry ramp; live dawn VPD 0.82 and dew margin 12.1F. Prior overnight plan was safe but forecast dryness was overstated.”},“stress_windows”:[{“kind”:“vpd_high”,“start”:“2026-09-11T10:00:00-06:00”,“end”:“2026-09-11T19:00:00-06:00”,“severity”:“high”,“mitigation”:“Live-gated vent-mist assist near band edge; engage excess 0.05, all-zone 0.25, fog excess 0.15.”},{“kind”:“heat”,“start”:“2026-09-11T13:00:00-06:00”,“end”:“2026-09-11T18:00:00-06:00”,“severity”:“high”,“mitigation”:“Pre-cool before peak, keep ventilation dominant, and accept capacity-limited heat.”}],“rationale”:[{“parameter”:“mister_engage_kpa”,“old_value”:1.6,“new_value”:1.3,“forecast_anchor”:“Corrected VPD rises above 2.7 kPa from late morning through 18:00… Result: Window scorecard: 77.1 planner score, 70.1% temperature compliance, 81.8% VPD compliance, 0.93 h total stress, 0.13 h VPD-high stress, 0 h VPD-low and 0 h dew-risk. Overnight recovery was safe, but temperature compliance remained the limiting axis and the hot/dry transition plan was only partly realized. Context also reports partial/missed transition execution and unavailable resource terms. Score: 6/10

New finding: For overnight-to-day plans, use live measured-edge recovery and keep the plan compact; forecast-driven moisture posture is useful only when transition writes are actually applied, so treat missed transitions as execution risk and verify persistence/readback before relying on them. → Added to Lessons Learned

Hypothesis

Setpoints

Saturday September 12

Primary crop-band changes:

Tactical tunable changes:

Changed secondary parameters:

🌅 Morning Cycle (6:42 AM) — iris-20260912-0640

Reflection

Validating previous cycle: iris-20260912-0017

Previous hypothesis: Prior plan iris-20260911-0639 scored 7/10 vs anchor 8: its 79.6 window score had 72.4% temperature compliance, 78.0% VPD compliance, 1.81 h stress, 0.57 h VPD-high exposure, and 0 h dew-risk. Safety held, but transition execution was partial. At midnight the house is 68.6F, VPD 0.75 kPa, IDLE, with 10.7F dew margin. Use measured-edge overnight recovery, then live-gated moisture assist through the morning ramp. Forecast-corrected conditions imply a cool pre-dawn followed by a hot, dry afternoon; precondition modestly without forecast-only wetting, preserve an 8F dew floor, and make assist available before midday VPD rises. Expected outcome: 0 h dew-risk and VPD-low overshoot, stable pre-dawn operation, and lower controller-attributable VPD-high exposure than 09-11 while accepting capacity-limited heat. Historical retrieval was unavailable because the embedding credential is missing. Result: Window scorecard for iris-20260912-0640: planner score 79.0, temperature compliance 75.6%, VPD compliance 81.1%, 1.65 h total stress including 0.23 h VPD-high, 0 h dew-risk; resource terms unavailable. Structured forecast missed the realized conditions: actual outdoor peak 77.5F vs 85.0F predicted, indoor-proxy RH minimum 47% vs 21%, and solar peak 1021 vs 495 W/m2. The expected hot/dry and heat-stress window did not occur; the moisture posture reduced risk but transition execution was partial, with multiple plan values reported missed or held by dispatcher audit. Hypothesis was only partly met: safety and dew margin held, but compliance remained moderate and the forecast anchor was poorly calibrated. Score: 6/10

New finding: When a hot/dry plan is driven by a materially over-dry temperature/VPD forecast, keep moisture escalation forecast-aware but require live VPD confirmation before aggressive posture; partial transition execution can dominate outcome, so verify waypoint persistence/readback and score execution fidelity separately. → Added to Lessons Learned

Hypothesis

Testing: Test band-coupled daytime wet assist with early engagement and 22s pulse gap, then taper only after measured VPD recovery. Expected outcome: Zero dew-risk and VPD-low hours; reduce controller-attributable VPD-high exposure below yesterday’s 1.20h while accepting capacity-limited heat.

Conditions

Cool dawn then hot-dry afternoon; prior overnight plan was safe but self-score 8 versus anchor 9 because VPD compliance was 84% despite zero stress. Retrieval was unavailable due to missing embedding credentials.

Expected stress windows

Parameter rationale

Setpoints

Saturday September 12

Primary crop-band changes:

Tactical tunable changes:

Changed secondary parameters:

🌆 Evening Cycle (7:17 PM) — iris-20260912-1915

Reflection

Validating previous cycle: iris-20260912-0640

Previous hypothesis: {“conditions”:{“outdoor_temp_peak_f”:85.0,“outdoor_rh_min_pct”:21.0,“solar_peak_w_m2”:495.0,“cloud_cover_avg_pct”:55.0,“notes”:“Cool dawn then hot-dry afternoon; prior overnight plan was safe but self-score 8 versus anchor 9 because VPD compliance was 84% despite zero stress. Retrieval was unavailable due to missing embedding credentials.”},“stress_windows”:[{“kind”:“vpd_high”,“start”:“2026-09-12T10:00:00-06:00”,“end”:“2026-09-12T19:00:00-06:00”,“severity”:“high”,“mitigation”:“Keep band-coupled assist near the high edge; fog excess 0.15-0.20 only while firmware gates permit.”},{“kind”:“heat”,“start”:“2026-09-12T14:00:00-06:00”,“end”:“2026-09-12T18:30:00-06:00”,“severity”:“high”,“mitigation”:“Precondition before peak, ventilate, and accept capacity-limited heat above the served envelope.”}],“rationale”:[{“parameter”:“mister_engage_kpa”,“old_value”:1.6,“new_value”:1… Result: Window score 79.1; temp compliance 82.3%, VPD compliance 76.2%; 0.00 h total, heat, VPD-high, VPD-low, and dew-risk stress. Water was not available for the window. Safety held, but VPD compliance lagged and house-too-wet drift was later observed at 0.59-0.64 kPa; the conservative overnight posture protected dew margin without fully centering the Vanda corridor. No durable new lesson beyond tapering only after measured recovery. Score: 6/10

Hypothesis

Conditions

Clear cool night: outdoor 71F to 53F, forecast VPD 1.89 to 0.39; live house 76.9F, VPD 0.79, dew margin 8.5F.

Expected stress windows

Parameter rationale

Setpoints

Saturday September 12

Primary crop-band changes:

Tactical tunable changes:

Sunday September 13

Primary crop-band changes:

Tactical tunable changes:

Changed secondary parameters:

End-of-Day Summary

Climate

Interior Light Evidence

• Crop DLI: unavailable
• Reason: source_reading_missing
• Provenance: legacy_invalid_exterior_proxy_plus_fixture_estimate
• Validity: dli-validity-v1; 2024-01-01T00:00:00+00:00 → open
• Qualified-light minutes and photoperiod remain independent of unavailable DLI.

Stress Hours

• Heat stress (>85°F): 2.5h
• VPD stress (>2.0 kPa): 8.4h
• Cold stress (<55°F): 12.1h

Economics

Equipment Runtimes

Water

• Total: 75 gal
• Mister: 23 gal

Hourly Pattern

7-Day Stress Context