Production Scheduling Glossary

Sixty-some terms that come up constantly in production scheduling — whether you’re evaluating an APS, reading the rest of this Learn Center, or just trying to make sense of a Microsoft Dynamics 365 Business Central manufacturing screen. Each entry links to a full article where one exists.

Advanced Planning and Scheduling (APS)

A category of software that sits above basic MRP and adds capacity-aware, sequence-aware scheduling — turning dated order suggestions into an executable plan that respects what machines and people can actually do at once. See What Is an APS?

Available-to-Promise (ATP)

A promise date based on what’s already in inventory or already on order — it answers “do I have (or will I have) the stock,” without considering whether there’s capacity free to actually produce more. Compare Capable-to-Promise.

Backward scheduling

Starting from a due date and working backward through a routing to find a starting date. If the routing needs more time than exists before the due date, the computed start lands in the past — an honest signal the date can’t be met as planned, not a system error. See Forward vs Backward Scheduling.

Bill of Materials (BOM)

The list of components and quantities consumed to produce one unit of a manufactured item. A BOM answers “what does this take”; a routing answers “how is it made” — together they drive a production order’s material and capacity needs.

Bottleneck

The resource whose limited capacity caps how much the whole system can produce, regardless of how much spare capacity exists everywhere else. An hour lost at a genuine bottleneck is an hour the whole plant loses; an hour lost anywhere else usually isn’t. See Bottlenecks and Theory of Constraints.

Calendar entry

The concrete record a shop calendar turns into once it’s assigned to a work or machine center and calculated: one row per resource per working period, carrying an actual date, a start and end time, and an effective capacity figure. A shop calendar is the reusable weekly pattern; calendar entries are the exploded, resource-specific rows that a scheduling calculation actually reads. See Work Centers, Machine Centers, and Shop Calendars.

Capable-to-Promise (CTP)

A promise date based on actually solving a schedule against real capacity and material availability — a genuine “can we make this by then” answer, rather than a lookup against existing stock. See What-If Scheduling and Capable-to-Promise.

Capacity

How much work a resource can perform in a given period — a function of its shop calendar, how many units run in parallel, and its efficiency percentage. Capacity is supply; compare it against demand (load) to see where a resource is overloaded. See Work Centers, Machine Centers, and Shop Calendars.

Capacity Constrained Resource (CCR)

Business Central’s one built-in exception to infinite loading: a work or machine center you explicitly register so the scheduler avoids stacking orders on top of each other’s already-committed time there. It’s a single-resource patch, not shop-wide finite scheduling — no sequencing logic, off by default. See Business Central Is Infinite-Loading.

Capacity leveling

Smoothing demand across a resource’s available time so it isn’t wildly overloaded in one period and idle in the next — a softer goal than strict finite scheduling, which prevents overload outright rather than just spreading it out.

Changeover

The time and effort needed to reconfigure a resource between two jobs — a tool swap, a color flush, a temperature ramp. When that time depends on which job ran before it, it’s sequence-dependent, and the order jobs run in starts to matter as much as which jobs run at all. See Sequence-Dependent Setup.

Constraint

A rule a schedule must (or should) respect — a machine that can’t run two jobs at once, an operation that can’t start before its predecessor finishes, a due date. Constraint-based scheduling builds a shop’s rules directly into the math rather than working around them with heuristics. See Constraint-Based Scheduling.

Constraint Programming (CP)

A mathematical modeling approach where a problem is expressed as variables and constraints, then solved by a search that either finds an assignment satisfying every constraint or proves none exists. A constraint solver of this kind can both find schedules and prove how close to optimal they are. See Constraint-Based Scheduling.

Cumulative constraint

A scheduling rule stating that at any moment, the total demand on a resource with more than one unit of capacity (three parallel machines, say) can’t exceed how many units actually exist. It generalizes a simpler rule — no two jobs on one machine at once — to resources that can run several jobs in parallel. See Constraint-Based Scheduling.

Dispatch list

A simple, priority-ordered list of what to run next on a resource — usually generated by a dispatch rule rather than a full schedule solve. It tells you what’s next, but not exactly when things start or finish, or whether the list as a whole actually fits the available time.

Dispatch rule

A quick, rule-of-thumb way to decide what runs next on a resource — earliest due date first, shortest processing time first, and similar heuristics. Fast to compute and easy to explain, but myopic: each decision looks only at what’s queued right now, with no view of the whole schedule. See Scheduling Algorithms Explained.

Drum-Buffer-Rope

A Theory of Constraints scheduling method that paces the entire plant to the bottleneck’s rate (the “drum”), protects it with a time buffer against upstream disruption, and limits how much work is released elsewhere (the “rope”) so work-in-process doesn’t pile up ahead of it. See Bottlenecks and Theory of Constraints.

Due date

The date a customer or downstream process needs an order by. Distinct from a promise date, which is what you actually communicate back after checking whether the due date is realistic against real capacity.

Efficiency %

The percentage of expected standard output a resource actually delivers — 100% means real output matches book time exactly; a lower figure means the real world is slower than the standard assumes. Efficiency scales an operation’s duration without changing the “book” time recorded on the routing.

Finite loading

Scheduling that respects a resource’s capacity as a hard ceiling for a specific resource or time window — refusing to place more demand on it than it can actually absorb in that window. Narrower than full finite scheduling, which reasons about an entire shop’s resources and sequence together. See Finite vs Infinite Capacity Scheduling.

Finite scheduling

Producing a complete, executable plan across every modeled resource where no resource is ever asked to do more than it can handle at once — the shop-wide version of finite loading, including a genuine decision about sequence and priority when multiple orders compete for the same time. See Finite vs Infinite Capacity Scheduling.

Firm planned order

A production order status one step more durable than “planned” — it’s survived being turned from a planning suggestion into a real intention, and it won’t be deleted the next time MRP runs, though it hasn’t yet been released to consume material or capacity for real. See Production Order Statuses in BC.

Flow shop

A production layout where every job follows the same sequence of resources in the same order, like an assembly line. Scheduling a flow shop is a more constrained (and often more tractable) problem than a general job shop, because the routing structure itself is identical for every order.

Forward scheduling

Starting from a starting date and working forward through a routing to find a finishing date. Its failure mode is the mirror of backward scheduling’s: it can compute a finish date well past when the order is actually needed, with nothing forcing the schedule to compress toward the due date. See Forward vs Backward Scheduling.

Frozen horizon

A near-term window, measured from today, inside which a schedule is treated as untouchable — protecting what’s already been communicated to the floor from being casually rewritten every time the plan re-solves. See Frozen Horizon and Plan Stability.

Gantt chart

A bar-chart view of a schedule, with resources (or orders) on one axis and time on the other, showing when each operation starts and ends. It’s the standard way to see and interact with a production schedule visually, rather than as a table of dates.

Heuristic

A method that finds a good answer quickly using rules of thumb, without guaranteeing the answer is the best possible one, or even knowing how close it is. The tradeoff against an exact solver is speed and simplicity versus a provable quality guarantee. See Scheduling Algorithms Explained.

Infinite loading

Scheduling that respects a resource’s calendar (its open hours) but not its existing commitments — placing every order as though it were the only thing using that resource, so multiple orders can land on the same machine at the same time without anything flagging it as invalid until after the fact. See Finite vs Infinite Capacity Scheduling.

Job shop

A production layout where different jobs can follow different sequences of resources, because different products need different operations in different orders. Scheduling a general job shop is a harder, more flexible problem than scheduling a flow shop, because there’s no single fixed routing shared by every order.

Lead time

The total elapsed time from when work on an order could start to when it’s finished — the sum of every operation’s setup, run, wait, and move time, plus queue time at each resource along the way. See Setup, Run, Wait, Move, Queue.

Lot splitting

Dividing a large operation into smaller sub-lots that can run in parallel across multiple machines, or across days, so the whole quantity finishes sooner than running it as one unbroken batch would allow. See Lot Splitting, Overlapping, and Transfer Batches.

Machine center

A finer-grained capacity object beneath a work center, representing one specific physical machine or resource rather than an interchangeable group. Optional — many shops model capacity at the work center level alone. See Work Centers, Machine Centers, and Shop Calendars.

Makespan

The total elapsed time from the start of the first operation in a schedule to the finish of the last one. Minimizing makespan means finishing the whole batch of work as early as possible — a different goal from minimizing tardiness, which cares about individual due dates rather than the overall finish. See Measuring a Schedule.

Manufacturing Execution System (MES)

Software that tracks and directs execution on the shop floor in real time — recording actual start/stop times, quantities, and quality data as work happens. An MES executes and records a plan; an APS is what decides the plan in the first place.

Move time

The transit time to physically get a part from one work center to the next. It extends an order’s elapsed schedule without occupying either work center’s capacity while it happens. See Setup, Run, Wait, Move, Queue.

Master Production Schedule (MPS)

The calculation of what to produce and when for end items with direct demand — sales orders and forecasts — one layer of the broader planning calculation that also includes MRP for dependent, component-level demand. See The Planning Worksheet Plans. It Doesn’t Schedule.

Material Requirements Planning (MRP)

A calculation that nets demand against supply and inventory, then suggests replenishment orders (make or buy) sized and dated to close the gap. MRP produces dated quantities; it doesn’t decide the sequence multiple orders run in on a shared resource. See Why MRP Is Not a Schedule.

MRP II

Manufacturing Resource Planning — the historical extension of MRP to also account for capacity, not just material, marking an early step toward what modern APS tools do more completely with exact, provable scheduling methods.

Nervousness

The tendency of a replanned schedule to change substantially in response to a small input change, even when the new plan is arguably “better” by the numbers — a real operational cost, because every change means reprinting instructions and re-coordinating a shop floor that already committed to the old plan. See Frozen Horizon and Plan Stability.

No-overlap constraint

The rule that a single-capacity resource can run only one operation at a time — the most basic constraint a finite scheduler enforces, and the one infinite loading never checks by default. See Constraint-Based Scheduling.

NP-hard

A computational-complexity classification meaning no known method can guarantee the exact best answer to every instance of the problem in a reasonable amount of time as the problem grows large. Most realistic job-shop scheduling problems are NP-hard, which is why exact solvers use time limits and report provable bounds rather than promising perfection instantly. See Scheduling Algorithms Explained.

On-time delivery

The simple measure of whether an order shipped by its due date, expressed as a percentage across all orders in a period. See On-Time-In-Full for the stricter version that also accounts for quantity.

On-Time-In-Full (OTIF)

A stricter delivery metric than plain on-time delivery: an order only counts as a success if it shipped on time and with the full quantity ordered — partial or late shipments both count against it. See Measuring a Schedule.

Operation

One step in a routing — a single unit of work assigned to one work center or machine center, with its own setup, run, wait, and move times. A routing is an ordered sequence of operations.

Overlapping

Starting a downstream operation before the upstream operation has finished its entire quantity, by sending a partial lot ahead — reducing total lead time at the cost of extra coordination between the two operations. See Lot Splitting, Overlapping, and Transfer Batches.

Pinning

Marking a specific operation’s scheduled time as something a solver should treat as fixed (or heavily preferred) rather than free to move, typically because it’s already been communicated to the floor or is already in progress.

Planned order

A production order status generated automatically by a planning run, carrying planning’s best current estimate of quantity and dates — and disposable, in that the next planning run regenerates it from scratch rather than preserving hand edits. See Production Order Statuses in BC.

Planning horizon

The forward-looking window of time a planning or scheduling calculation actually considers — orders and capacity beyond that window simply aren’t part of the current calculation, whether or not they exist in the system.

Precedence

The rule that one operation in a routing must finish before the next one in sequence can start. A finite scheduler treats precedence as a hard rule while it’s also respecting capacity — the schedule has to be valid on both fronts simultaneously, not just one at a time.

Production order

The core manufacturing document in Business Central, carrying a routing (operations) and components (materials) for converting purchased or produced material into a finished item, and moving through a defined sequence of statuses over its life. See Production Order Statuses in BC.

Promise date

The date actually communicated back to a customer, after checking real availability and capacity — distinct from a due date, which is what was originally requested and may or may not be achievable.

Queue time

A fixed buffer set on a work center, representing the gap between when material physically arrives and when work on it can actually start. It’s set on the resource, not the operation, and adds to lead time without itself being setup or run work. See Setup, Run, Wait, Move, Queue.

Released order

A production order status where work is actually authorized to begin — materials can be consumed and output posted against it, using whichever flushing method the item and work center define. See Production Order Statuses in BC.

Routing

The ordered list of operations, and which work or machine center each one uses, that defines how an item is actually produced. Alongside the BOM (what’s consumed), the routing is the master data a production order’s schedule is built from. See How Business Central Schedules a Production Order.

Run time

The per-unit processing time for an operation, multiplied by quantity — the one routing time element that scales directly with lot size. See Setup, Run, Wait, Move, Queue.

Schedule adherence

How closely an actual (or newly re-solved) schedule matches a previously published one — a KPI that captures plan stability, distinct from and sometimes in tension with pure optimality. See Measuring a Schedule.

Sequence-dependent setup

Changeover time that depends on which job ran immediately before the current one, not just on the current job alone — modeled with a from-to matrix rather than a single flat setup value per item. See Sequence-Dependent Setup.

Setup time

The one-time preparation an operation needs before work can start — a tool change, a calibration — charged once per operation regardless of how many units run. See Setup, Run, Wait, Move, Queue.

Shop calendar

The master pattern defining a resource’s standard working week — shifts, start/end times, and holidays — which gets calculated into concrete calendar entries once assigned to a work center. See Work Centers, Machine Centers, and Shop Calendars.

Simulated order

A production order status used purely for costing or quoting estimates — it’s invisible to planning and carries no capacity load, and its underlying record disappears the moment its status changes. See Production Order Statuses in BC.

Tardiness

How late an order finishes relative to its due date — zero if it’s on time or early, and a positive number growing with lateness otherwise. Total, maximum, and weighted tardiness are all common variants a schedule can be optimized against. See Measuring a Schedule.

Theory of Constraints (TOC)

A management framework, associated with Eliyahu Goldratt, built around the idea that a system’s overall output is governed by its single tightest constraint — so improving anything that isn’t the current bottleneck doesn’t improve the whole system’s throughput. See Bottlenecks and Theory of Constraints.

Transfer batch

The quantity moved from one operation to the next before the entire lot at the upstream operation is finished — the mechanism that makes overlapping possible. See Lot Splitting, Overlapping, and Transfer Batches.

Utilization

The percentage of a resource’s available capacity that’s actually in use. Counterintuitively, a healthy shop usually doesn’t target 100%: some slack capacity absorbs variability without every hiccup cascading into missed due dates elsewhere. See Measuring a Schedule.

Wait time

Time that has to pass after an operation’s run finishes before the part can move on — drying, cooling, curing. Like move time, it extends the elapsed schedule without occupying the work center’s own capacity. See Setup, Run, Wait, Move, Queue.

What-if scenario

A hypothetical alternative to the live schedule — add a shift, expedite an order, add a machine — solved and compared against the current baseline before anything is actually committed. See What-If Scheduling and Capable-to-Promise.

Work In Process (WIP)

Orders and materials that are actively being worked on but not yet finished — released production orders that have started consuming material and capacity but haven’t posted final output.

Work center

The core capacity object every routing operation ultimately schedules against — a department, resource, or group of interchangeable capacity, carrying its own calendar, capacity, and efficiency. See Work Centers, Machine Centers, and Shop Calendars.