Skip to content

Implementing scheduling

Kinetic + Classic

The scheduling engine is only as good as the data behind it. Most “scheduling doesn’t work” complaints come down to calendars, resources or routings that don’t describe the real shop. This checklist covers what has to be true before the schedule is worth trusting, roughly in the order to tackle it.

  • Decide on the default scheduling direction (backward from the required date, or forward from today).
  • On the site’s planning settings, set the Finite Horizon, Rough Cut Horizon, Overload Horizon and Scheduling Send Ahead For option.
  • Set up Scheduling Priority Codes if you want job priorities to influence the global order.
  • Create Production Calendars for the company, each site and any resource group or resource that works different hours. Add holidays, planned shutdowns and overtime as calendar exceptions.
  • Use Shifts for breaks rather than cutting gaps into a calendar day. A calendar day with a gap in the middle (for example four hours, a gap, then four hours) can make the engine loop.
  • Give suppliers calendars too if subcontract lead times depend on their working days.
  • Build Resource Groups that match real work centres, with a resource for each machine or person that can be scheduled separately.
  • On each, set the calendar, Finite Capacity, queue and move hours and, if needed, Split Operations.
  • Check the Number of Resources on each group matches the resources actually listed under it.
  • See Resources, resource groups and capabilities.
  • Every operation needs realistic setup hours and a production standard, and exactly one scheduling resource: a resource group, a resource or a capability.
  • Include subcontract operations with realistic days out.
  • Clean up both the part methods and the open jobs. Open jobs carry their own copy of the method, so fixing Engineering Workbench alone doesn’t fix what’s already scheduled.
  • Very large production hours on a single operation (over about 100 hours) are worth checking; they are a known cause of scheduling runs that hang.

Set part type, UOMs, manufacturing lot sizes, lead times, days of supply, Receive Time, minimum on-hand, safety stock and the MRP flags. These drive the jobs and POs that MRP creates, and so the load the scheduler sees.

Keep sales order, job and PO release dates current. The schedule can’t be better than the dates it’s aiming at. If demand changes daily, the changes have to be entered daily.

Put these in a Process Set scheduled at least once a day, overnight if possible:

  1. Process MRP (job and PO suggestions)
  2. Calculate Global Scheduling Order
  3. Global Scheduling
  4. Any load-graph or cube refresh processes you use

A one-time cleanup run may be needed at go-live to clear arrears and rebuild load.

The schedule updates from labor. Operators need to clock the right operation, report quantities and downtime accurately, and tick Complete when an operation is done. Operations left open or reported out of sequence keep load on resources that are actually free.

Screen Use it to
Job Scheduling Board See and drag a job’s operations; lock jobs so rescheduling leaves them alone
Resource Scheduling Board See what’s loaded on each resource and move operations between resources
Resource Schedule Load Graph, Site Schedule Load Graph Compare load with capacity over time
Overload Informer List resources that are over capacity within the overload horizon
Work Queue Show operators what’s next on their resource
Shop Load, Priority Dispatch reports Printed load and dispatch lists
Schedule Impact report See what a change would push out
Projected Sales Order Shortages Find orders the current schedule will miss
Master Production Schedule Plan finished-goods output

Sources

This page is written in our own words. It was informed by: