Skip to main content
Beta
Coming Soon. Job-level scheduling — assigning jobs to specific resources with detailed timelines, changeover-aware sequencing, and downtime-aware feasibility — is on the roadmap and is not available in the product today. This page describes what the capability will cover and what to use in the meantime.

​
Overview

Production Planning answers what to produce and when, at the granularity of the planning horizon’s interval — typically a day, a week, or a month. Job scheduling will answer who does what, on which resource, in what order, at the granularity of individual jobs with start and end times: a floor-level plan you could hand to a shift lead. The two are complementary, not substitutes. Planning commits to production quantities per interval; scheduling turns those quantities into a sequenced timeline on specific resources.
Two things in the product already carry the word “schedule,” and neither is job scheduling: simulation schedules are timed inputs to a simulation model (when entities are released, when scheduled actions fire), and scheduled jobs are recurring Dexter tasks that run on a timer.

​
What Today’s Product Covers

Two capabilities cover parts of the problem job scheduling will eventually solve end-to-end:
  • Production Planning decides interval-level production quantities against demand, on-hand inventory, resource capacity, and inventory goals. It respects hard capacity limits and honors deadlines through order tags, which can be soft (a weighted lateness penalty) or hard (an enforced constraint). If your question is “given this demand book and this capacity, what should we produce next week?” — planning is the answer.
  • Simulation models the floor as-is — resources, process flow, shift patterns, downtime, variability — and plays out a given production plan or arrival stream to see what actually happens. If your question is “will the floor actually hit these numbers given shifts, changeovers, and typical variability?” — simulation is the answer.
Together they cover the two ends: planning gives you the quantities to commit to, and simulation pressure-tests whether the floor can execute them. Simulation is descriptive rather than prescriptive, though — it does not choose the job sequence, it plays out one you give it.

​
The Gap Job Scheduling Will Close

Between “produce 4,200 units of SKU-A next week” and “run job #47 on Line A starting Tuesday 09:12” sits the sequencing problem. That is what job scheduling will address when it ships:
  • Assigning jobs to specific resources — not “the packaging pool has 80 hours this week,” but “this job runs on Packer 2”
  • Changeover-aware sequencing — a changeover matrix (the setup-time cost of switching from one job type to another on a given resource) drives an order that keeps setup time low
  • Per-resource downtime and availability calendars — maintenance windows, shift patterns, and holidays that block job placement on a specific machine
  • Job-level precedence from the BOM graph — a component’s job must finish before its parent’s job starts, read from the same BOM structure planning uses
  • A job-level timeline as output — a Gantt chart of start and end times per job per resource, changeover gaps visible, unfulfilled jobs flagged when capacity is insufficient
  • Sequencing KPIs — makespan, resource utilization, changeover time as a share of run time
  • Manual edits on top of an optimized schedule, with both versions retained
  • Scenario comparison across scheduling runs with different downtime calendars, capacities, or targets
None of these exist in the product today. This page will be expanded as the capability lands.

​
What to Do in the Meantime

If you need floor-level sequencing today, the practical path is:
  1. Use Production Planning to fix the interval-level production quantities against demand, capacity, and inventory goals.
  2. Use Simulation to model the floor and validate that the planned quantities survive realistic timing, downtime, and variability — simulation schedules let you specify when entities arrive and when actions fire.
  3. For the sequencing decisions themselves — job order on a specific machine, changeover-minimizing assignments — keep your existing scheduling process (scheduling board, MES, or spreadsheet) and ask Dexter to help reason about trade-offs, encode rules, and cross-check against planning output and simulation results.
A common working pattern: let planning commit weekly quantities, hand Dexter the week’s committed quantities plus your machine list and changeover rules, and have him draft a candidate sequence you refine by hand — then simulate the result to check it holds up.