Work order prioritisation

The short answer

Work order prioritisation is deciding, when several jobs want the same machine, which one runs first — and on which machine it runs. It matters because capacity is finite: a machine can only do one thing at a time, so the moment more work is loaded onto it than it has hours for, the sequence you choose is what decides which orders ship on time and which slip. Get it right and the plant hits its dates with the capacity it already has; get it wrong and you miss deliveries while machines sit idle in the wrong order.

There are two distinct decisions inside the word "priority". Order priority ranks the queue — which work order or customer order is more urgent. Resource priority routes it — which machine or work centre should do the operation when more than one can. On top of those sits a set of sequencing rules — earliest due date, customer priority, bottleneck-first, grouping by setup — and a Gantt board for dragging the result into shape when a machine tips over 100% loading. This guide works through each, grounded in how production planning software actually does it.

Why sequence matters at all

If every machine had infinite capacity, prioritisation would be pointless — every job could start the instant it was ready, and nothing would ever wait. Real shop floors are the opposite. A CNC machine, a heat-treatment furnace, an assembly cell: each can process one job at a time, and the total work asked of it in a week routinely exceeds the hours in that week. That mismatch — more demand than capacity on the constraint — is the entire reason sequencing exists.

Once a machine is over-committed, order matters in a way it never does on paper. Two work orders due the same Friday cannot both run first. Whichever you sequence ahead ships on time; the other slips unless you find capacity elsewhere. Multiply that across a plant of shared machines and the sequence becomes the single biggest lever a planner has over on-time delivery — bigger, usually, than buying another machine. The job of a planning system is to make that sequence visible and adjustable instead of leaving it in a supervisor's head, where it is invisible until a job is already late.

Order priority vs resource priority

The word "priority" hides two different questions. Confusing them is why so many shops set a priority number and still miss dates — they ranked the queue but never fixed where the jobs run, or vice versa.

In Fast Planning, order priority is explicit: OrderPriority.aspx sets it and ChangeOrderPriority.aspx re-sequences it when the picture changes — a rush order lands, or a delayed part pushes a job back. Resource priority is carried on the work-order routing and made visible on the machine loading and scheduling and priority (Gantt) boards, so a planner can see where each operation is headed and move it if that machine is already drowning. You need both working together: order priority ranks the queue, resource priority spreads it across the machines that can absorb it.

Finite machine loading — the constraint

Prioritisation only becomes necessary because loading is finite. The measure is simple. For any machine over a period, percentage loading = load hours ÷ available hours, where the load a work order puts on a machine is standard cycle time × quantity + setting time. Add up the load of every operation routed to that machine, divide by the hours it actually has after shifts, breaks and planned maintenance, and you get its % loading for the day or week.

Below 100%, there is slack — the sequence is comfortable and most rules will hit their dates. At or above 100%, the machine is a constraint, and the sequence stops being a convenience and becomes the thing that decides who is late. A machine loaded to 130% for the week is telling you plainly that 30% of what you asked of it will not happen on time unless you re-sequence, offload, or add capacity. Infinite-capacity planning hides this by pretending every job can start at once; finite machine loading surfaces it, which is exactly what makes prioritisation a real decision rather than a formality.

The sequencing rules that decide first

Once you accept that something must run first, the question is which rule chooses. No single rule wins every time, so experienced planners blend a handful — and understanding what each optimises for is what lets you blend them well.

• Earliest due date (EDD). Sequence by the promised date — the job due soonest runs first. It is the natural default because it directly protects on-time delivery, and it is the baseline most schedules fall back to.
• Customer / order priority. A strategic customer or a rush order overrides EDD and jumps the queue. This is the manual lever — set in OrderPriority.aspx — for when the business reason outweighs the calendar.
• Bottleneck-first. Sequence the constraint machine before anything else, because the bottleneck sets the plant's whole output. An hour lost on the bottleneck is an hour lost for the factory; an hour lost on a non-bottleneck is usually free.
• Group by setup. Batch jobs that share tooling or a fixture so you change over once instead of five times. Setting time is dead capacity, and grouping wins it back — but only where it does not push a genuine due date past its date.

A workable order of operations is: protect the bottleneck, honour real customer priority, fall back to earliest due date for everything else, and group by setup wherever it does not break a date. The steps below show how that decision runs end to end, from a raised work order to a printed job card on the floor.

The final step is not an afterthought. Once the sequence is fixed, job cards are printed so the floor works to it — individually through WOSpecifPrintOrder.aspx when a single order is released, or in batch through multiplewojobprint.aspx when a whole shift's worth of work orders goes out at once. Each card carries the WO specification and routing (clsWOSpecification) that lists the operations, so the sequence on the board and the paperwork on the machine are the same sequence.

The Gantt board and re-sequencing

Rules propose a sequence; a Gantt board is where you check it against reality and fix it. Fast Planning uses a DayPilot scheduler board (DayPilot.dll) that lays work orders out as bars across a timeline, one row per machine, so overloads and clashes are visible at a glance instead of buried in a list. A planner drags a bar to move it earlier or later, or onto a different machine, and the board reflows — the same drag-oriented scheduling the platform uses for project Gantts.

The board earns its keep when a machine crosses 100%. There are only three honest ways to clear an overload, and the Gantt lets you test each before committing:

Whichever lever you pull, the point of doing it on the board is that you see the result before it reaches the floor. And because the load is finite, the schedule can then offer a projected availability — the realistic next free slot on each machine — so sales promises a date the shop can actually keep rather than an optimistic guess that becomes tomorrow's late order. Re-sequencing in ChangeOrderPriority.aspx and dragging on the Gantt are two views of the same act: deciding, honestly and against real capacity, what runs first.

Job-work, OSL and the India reality

For Indian machine shops and component makers, two realities sit on top of the textbook sequencing rules. The first is job-work and outsourced (OSL) operations in the routing. A large share of work leaves the gate mid-process — for plating, heat treatment, grinding or specialist machining — and that leg adds its own lead time to the sequence. The in-house operations that follow cannot start until the parts come back, so the schedule has to treat a send-out-and-return the same way it treats machine time. An over-committed job-worker misses a due date exactly as an overloaded machine does, and a sequence that ignores the OSL leg promises dates it cannot keep. Carrying the outsourced operation on the work-order routing is what lets the board show the real lead time instead of assuming everything happens under one roof.

The second is the machine-shop bottleneck reality: in a typical small-to-mid Indian unit, one or two expensive machines — a VMC, a specific grinder — are the constraint the whole plant lives around, and sequencing them well is worth more than any other planning effort. Getting that right with the capacity already on the floor is usually cheaper than adding a machine. On the commercials, indicative INR pricing for the planning module is on the pricing page; treat those figures as a starting point and confirm the numbers and any tax treatment with your CA, since a planning schedule is an operations tool, not tax advice.

How Fast Planning prioritises work orders

Fast Planning Software — built by Improsys in Pune under the Fast Technology brand, cloud or on-premise — turns prioritisation from a whiteboard habit into a scheduled decision, because the priority, the load and the routing are all the same records.

Frequently asked questions