Line scheduling modes: six ways to run a production line
Last updated: July 22, 2026 · 8 min read
A production line is a group of machines product flows through in order. How tightly those machines are coupled is a physical fact about your equipment, so it should be a setting, not an assumption. There are six line scheduling modes, from fully independent machines to one continuous stream, and you set the mode per line to match how it actually runs.
First, how the scheduler thinks.
An order is a job with a due date. It moves through steps, and each step needs a machine that can run it. The scheduler places every step on a capable machine and packs the sequence against that machine's working hours. Nothing more exotic than that.
The duration for each step comes from your own run history, not a planner's guess. How long Step 1 actually takes on Machine 1 is a fact your floor already produces. The scheduler uses that fact.
On a job shop that is enough. Machines stand alone, and each one gets scheduled on its own. But many operations run production lines, where machines are physically connected and product flows through as a unit. Scheduling a line one machine at a time misses what actually governs it. So the scheduler can treat the line itself as the thing being scheduled.
Three schedules. Each machine has its own queue. A job goes to whichever capable machine is open, and the machines never wait on each other.
One schedule. The same three machines, scheduled together. Product enters and flows Machine 1 to 2 to 3 as a unit, and the line is what gets booked.
Every way to run a line.
A line is a group of machines that product flows through in order. How tightly they're coupled is a physical fact about your equipment, so it's a setting, not an assumption. There's a whole spectrum, from machines that stand alone to a line that runs as one continuous stream. Six behaviors cover it.
Each mode below follows two orders, A and B, through the same three-machine line, loosest coupling first. Faint blocks are changeover, the setup time before an order can run.
Independent
Every machine schedules on its own. Order B can start on Machine 1 while Order A still runs Machine 2. Throughput is tight because no machine waits on another. What you give up is coordination: the line isn't one unit, so nothing protects the flow between machines.
Use it when machines aren't physically connected. Standard job-shop behavior.
Pipelined
Order B follows Order A through the line, staggered one machine behind. As soon as A clears Machine 1 and it changes over, B starts on Machine 1 while A is still on Machine 2. There's no full lockout, so you keep most of independent scheduling's throughput while the line holds its order. An optional buffer sets a minimum gap between orders per machine when your process needs it.
Use it when the line runs continuously and orders should follow each other through without a full stop between them.
Partial Pipeline
The line stays reserved only up to a release point you choose, marked with a star. Order A holds Machine 1 and 2 until it clears that step, then the front frees and Order B starts while A finishes downstream. It splits the difference: the committed front is protected, the tail opens early for throughput.
Use it when the risk lives at the front of the line and the back end can safely take on the next order early. The release point is a setting on the step that frees it.
Full Reservation
The line is one atomic resource. Order A holds every machine until it clears the last, then changes over and Order B begins. The gap between orders is set by the slowest machine to change over, the bottleneck. It's the simplest honest model for a line where product is physically committed once it enters. The trade is throughput: Machine 3 sits idle while A is up front, and nothing else can use it.
Use it when product can't be interleaved once it enters. Changeover is the max of all machines when operators set up in parallel, or the sum when one operator works down the line.
Reserved + Setup Pre-Start
Full reservation with the dead time cut out. The line still runs one order at a time, but once a machine finishes its part of Order A it sits idle, so the scheduler starts Order B's setup there while A works downstream. When A clears the line, B's changeover is already done and it starts almost immediately. Same protection as full reservation, a shorter gap between orders.
Use it when a reserved line loses real time to changeover between orders and the upstream machines are sitting idle anyway.
Continuous Flow
Every machine works the same order at once. Product enters at Machine 1 and streams through all three as one connected flow, so its time on the line is roughly the longest single step, not the sum. Switching orders ripples through: Machine 1 changes over first while the tail of Order A still runs downstream. The gap is set by the bottleneck, the machine with the longest changeover, and setup can pre-start upstream to shrink it.
Use it when the equipment is physically one stream, conveyors and connected stages, with no meaningful notion of a part sitting between machines.
One model underneath all six: an order, its steps, and machines with working hours. What changes between modes is how much the line behaves as a single unit. That's yours to set per line, because it's a fact about your equipment, not the scheduler's to assume.
A plan is easy to draw at 6am. The test is 10am, when a machine goes down.
The plan changed. You should hear it from the schedule first.
A schedule that doesn't know what's happening on the floor is a guess by mid-morning. The scheduler reads live machine state, so it learns about trouble while there's still time to act on it. Three things make that real.
Some things don't move
A maintenance window at 2pm is an appointment, not a suggestion. You can pin a block to a clock time and it holds, even when the morning runs long and everything around it slides. Work flows past it. The technician shows up to an idle machine, the way it was promised. You choose what automation is allowed to touch and what it never moves.
It sees the slip before the parts do
Changeover is where days quietly go wrong. If setup on a machine runs past its planned time, the scheduler knows the order is behind before a single part comes off it. From there it projects the rest of the day forward and shows you the finish times you're actually heading toward, not the ones you planned.
When a machine goes down, it hands you the options
A machine drops offline. The scheduler marks it down, holds the window, and works out what happens to every job that was counting on it. It does not quietly rearrange your floor. It hands you the affected work sorted into three answers, and you approve the move.
Move it
Another machine can run this job and still hit the due date. The scheduler picks the one that adds the least changeover and makes nothing else late.
Wait for it
The machine comes back in time. A Friday order shouldn't churn across the floor for a two o'clock recovery, so waiting stays on the table as a real choice.
Call the customer
This one is late no matter what you do. It's flagged in red now, this morning, while you still have time to make the call yourself.
Every one of those moves previews before it applies and undoes cleanly if you change your mind. When the machine comes back, that's a new suggestion too, never a silent reversal of what you already decided.
The difference.
The usual Tuesday
You found out at 3pm that Line 2 had been behind since 9. The shift ended at 4. You had an hour and a customer relationship on the line, and the schedule on the wall stopped being true around mid-morning.
With the scheduler watching
At 9:20 the changeover ran long and the projected finish moved. You saw it, moved two jobs to a machine with room, and let one wait for recovery. The customer got a call before lunch, not an apology after.
The small things that make it usable.
Auto-schedule, scoped
Fill the whole board in one click, or just Thursday onward, or just the five orders you selected. Freshly placed work doesn't get rearranged by the same click that placed it.
Ripple with a preview
Insert a rush order, push a day out, or make room, and see exactly what moves and who runs late before you commit. Every action undoes.
Holidays and shifts
Close December 25, add a Saturday shift this week, extend today to 10pm. The calendar bends to the real one, and the plan bends with it.
Recurring maintenance
Set a cadence once and the windows appear weeks ahead, pinned in place so a busy morning never quietly runs them over.
Overnight and 24/7
Midnight stops being a seam. Work flows through it, changeovers straddle it, and the plan reads the way the floor actually runs.
One conflict list
Idle gaps, drift, unconfirmed downtime, late orders, capacity overrides. Everything that needs a decision, in one panel, each with a jump to it.
MACH Schedule ships all six modes. You pick one per line, and auto-scheduling honors it when it sequences orders across that group of machines. See line scheduling modes on the MACH Schedule page →
Line scheduling questions.
How does a scheduler run a connected production line?
When machines are physically connected as a line, the line itself can be scheduled as one resource instead of each machine on its own. How tightly the line is coupled is a physical fact about the equipment, so MACH treats it as a setting per line. That coupling runs across a spectrum, from fully independent machines to one continuous stream, so the plan matches how the line actually runs.
What are the line scheduling modes?
There are six line scheduling modes, arranged loosest coupling first: Independent (each machine schedules on its own), Pipelined (orders follow one machine behind), Partial Pipeline (the line frees at a release point you choose), Full Reservation (one order holds the whole line), Reserved + Setup Pre-Start (reserved, with the next order's setup overlapping to shrink the gap), and Continuous Flow (every machine runs the same order as one stream). You set the mode per line to match how the equipment is coupled.
What is changeover in line scheduling?
Changeover is the setup time on a machine before an order can run. In a reserved or continuous line, the gap between orders is set by the slowest machine to change over, the bottleneck. Some modes let the next order's setup pre-start on upstream machines that are already idle, which shrinks the gap between orders.