Production Planning

Export the open order book from NetSuite, Epicor Kinetic, or Dynamics 365 BC. Filter by promised ship date in the planning horizon (typically the next 4 weeks). Confirm each order has a current rev print and no holds.

Compare S&OP forecast to firm orders and kanban triggers. Flag any forecast bucket consumed early or under-consumed — those are the signals you'll level-load against in heijunka.

Compare promised ship date against standard lead time per routing. Any order with insufficient runway gets escalated to customer service same-day for re-promise or expedite.

Work the exception queue: reschedule-in, reschedule-out, expedite, cancel. Don't ignore reschedule-out messages — pulling them in clears WIP and frees cash.

Walk the floor on long-lead and high-value items rather than trusting ERP on-hand. Cycle-count variances commonly surface here. Capture lot/heat numbers required for traceability per AS9100 or IATF 16949.

Buyer contacts supplier for committed delivery date. Document the new promise on the PO. If the supplier can't meet need-by, escalate to second-source on the AVL or notify CS that the WO will slip.

Check tool crib for required tooling, dies, fixtures, and gauges. Verify gauge calibration is current — a red-tagged past-due gauge will halt FAI. Coordinate with toolroom on any sharpening or repair turnaround.

Compare planned hours per work center to available capacity (shifts × hours × OEE). Identify the constraint resource (CCR) — the schedule must protect the bottleneck per Theory of Constraints. Anything over 90% load is a flag.

Apply SMED logic — group by tooling family or material to minimize changeover. Don't sacrifice on-time delivery for setup efficiency on a CCR; protect customer due dates first.

Cross-check PLM (SolidWorks PDM, Arena, Windchill) for any pending or recently released ECN affecting open WOs. Releasing a job to an old print rev is a common scrap driver — and on regulated work (medical, aerospace) a deviation event.

Publish the dispatch list in Plex, Tulip, MachineMetrics, or your MES. Each work center gets a sequenced queue with target start/finish times.

Walk shift leads through the dispatch list, hot orders, known constraints, and any new ECN training required before run. Use the same gemba board the next shift's handover will use.

If an ECN released since last week, verify the assigned operators have signed the training acknowledgment. Running a new rev without trained operators is a common cause of first-piece rejects.

Pull OEE and throughput from MachineMetrics or MES. Identify variance drivers: downtime, scrap, slow cycles, missed setups. Variance under 5% is noise; over that needs a named cause.