Pre-Flight Gate (Before Commissioning)

Operational concept locked (flows, zones, wave logic, exceptions, safety rules).
Roles clarified: WMS vs WCS vs WES (who decides work, who controls machines, who orchestrates).
Interface inventory complete (WMS↔WCS/WES, ERP/OMS/TMS, labelers, scales, sortation, robotics).
Data standards set (item master, dimensions/weights, barcode rules, label formats, cartonization inputs).
Site readiness verified (power, network, RF coverage, safety signoff, spare parts plan).

Commissioning Quality Gate (Automation)

Mechanical checkout complete (conveyors, AS/RS, AMRs, sortation, scanners, printers).
Control logic validated (sensor mapping, jam handling, safe stop/restart procedures).
"Degraded mode" defined (how to operate manually when automation is down).
Maintenance readiness confirmed (PM schedules, spares, vendor response SLAs, training).

Go-live is not a single moment—automation needs a controlled ramp and rehearsals.

Message contracts validated (what fields, what timing, what acknowledgements).
Exception handling tested (timeouts, duplicates, replays, partial picks, short cartons).
Latency thresholds defined (what's "too slow" before it becomes a service issue).
Traceability validated end-to-end (carton/license plate tracking through automation).
Monitoring enabled (queues, error rates, jam counts, message failures, throughput).

Ramp plan defined (Day 1 volume cap, Day 2–14 increases, peak avoidance rules).
Order profiles rehearsed (small parcel, multi-line, bulky, fragile, returns).
Staffing plan aligned to ramp (super users, floor walkers, maintenance coverage).
Cutover rehearsal completed (timed runbook + go/no-go + rollback posture).
"Stop / slow down" triggers defined (safety, jam rates, backlog, mis-sorts, accuracy).

Happy-path demos aren't enough. Timeouts, duplicates, and partial confirmations are where startups fail.

If automation goes down, you still must ship. Define manual bypass workflows before Day 1.

Without a shared command center and clear responsibilities, WMS vs automation vendors will debate ownership while service suffers.

Single command center established (one intake, one priority list, one comms cadence).
Severity model + response SLAs agreed across vendors (WMS + automation + integrator).
Daily KPI review first (throughput, accuracy, jam rates, backlog, cycle time).
Fix validation rules defined (regression checks before deploying changes).
Exit criteria defined (no repeating Sev-1; stable KPIs for 5–10 business days).

Area	Org	WMS Vendor	Automation/WCS/WES	Shared Notes
L1 Support (floor issues, process questions)	R/A	C	C	Super users + floor walkers reduce noise.
Interface triage (message failures, retries)	R	R	R	Single triage meeting, shared logs.
Mechanical downtime and jams	C	C	R/A	Degraded mode + rapid recovery procedure.
Cutover / ramp decisions	A	R	R	Stop/go triggers must be respected.

Release calendar + simulation/regression test policy for interfaces.
Preventive maintenance governance (PM compliance, spares, failure modes tracked).
Monitoring ownership retained (queues, jams, error rates, throughput trends).
Quarterly "automation health" review (root causes, training gaps, uptime, capacity constraints).