Trip batching and site visit policy — two-site smart farm¶
Purpose¶
Operational policy for SITE_HOME ↔ SITE_FARM travel when COMMUTE_ONE_WAY ≈ 35 min. Batching, low daily touch, and remote observability are mandatory design filters—not nice-to-haves. Routine unnecessary trips are assumed to destroy margin and morale.
Canonical distance math: Two-site operating model — 5 ac / 120 ac (distance-tax rules).
Structural misfits: Two-site structure disqualifiers.
What justifies a trip¶
A visit is justified only if at least one applies before departure:
| Class | Justification | Examples (non-exhaustive) |
|---|---|---|
| J1 — Scheduled work order | Written list of tasks with minimum H_VISIT |
Fence repair block; working pens; loading; hay window work |
| J2 — Tier-1 alert | Open condition from Instrumentation priority matrix or documented life-safety / major asset risk |
Dry tank; pump fault; perimeter breach where verified |
| J3 — Surge calendar | Pre-declared surge (calving, shipping, storm recovery) | Overrides normal batch days |
| J4 — Third-party | Vet, buyer, inspector—clock bounded | Combine with J1 when possible |
Not a justification: “I’m worried,” “I haven’t been today,” “I’ll combine with errands” without J1–J4.
Junk trip = none of J1–J4 satisfied → counts toward policy failure (morale + review), not harmless.
What must be batched¶
Default: Stack all deferrable physical work into the fewest high-payload blocks per week.
| Must batch | Rationale |
|---|---|
| Routine animal moves / checks compatible with non-daily rhythm | T1 target—two-site operating model |
| Maintenance that is not time-critical | DT-1 minimum visit payload |
| Supply delivery (feed, parts) | Single truck run with list |
| Social / neighbor stops | Not farm trips—separate mental account |
Surge exception: When J3 is active, batching may mean “stay” or “return next day”—not tiny daily drops.
What should be remotely observable (first-class)¶
Priority order aligns with Instrumentation priority matrix:
| Priority | Observable | Why |
|---|---|---|
| P1 | Water level / flow / pump run/fault | Highest trip-kill per dollar |
| P2 | Power at critical loads (pump, gate opener) | Silent failures hurt |
| P3 | Gate position / latch state where legally and reliably done | Reduces “did I leave it” trips |
| P4 | Weather / soil moisture at decision points | Schedules when to batch field work—not daily scouting substitute |
Rule: If an observation does not change a decision or eliminate a trip, it is not P1—it is nice-to-have (Instrumentation ROI).
Triage cap: H_TRIAGE hours/week on alerts—hard ceiling per Enterprise unit economics. Overflow = de-scope sensors, not more driving.
Work patterns that indicate failure¶
| Signal | Interpretation |
|---|---|
trips_wk rises without herd/crop scale change |
Batching or telemetry failed—stop adding work |
| Rising junk trips (self-reported or calendar audit) | Morale + margin leak—enforce work orders |
| Alerts ignored after fatigue | Instrumentation design wrong—fewer channels, better rules |
| “Quick stop” visits multiplying | Distance tax denial—re-read DT-4 / disqualifiers |
| Primary operator cannot hold off-farm job and pattern | Enterprise or labor model wrong—not willpower |
| Second person always dragged for safety on routine tasks | Under-invested facilities or wrong enterprise intensity |
Escalation: Repeated failure signals → formal review against Execution gates — financial thresholds and Vision stop rules—freeze non-essential CAPEX.