Your schedule is fine on Monday and fiction by Wednesday
If you run a job shop, you know the feeling. Monday morning the schedule looks clean. Every machine has its jobs, every due date has a plan. By Wednesday afternoon it is a wall of scribbles, two jobs are late, and you are running the floor off your phone and your memory.
That is not because you are bad at planning. It is because you run high-mix, low-volume work, and HMLV breaks the tidy kind of schedule the same way every week. Understanding why is the first step to building a schedule that bends instead of shattering.
What makes high-mix, low-volume different
High-mix, low-volume means variety over quantity. Many part numbers, small batches, short runs. A shop that turns out 40 different parts a week, most in quantities of 5 to 50, lives here. Compare that to a high-volume plant running the same three parts for a month, and the difference is everything.
In a long production run, setup is a rounding error. You change over once and run the part 5,000 times, so the hours you spend dialing it in disappear against the hours of cutting. In HMLV, that math flips. You might spend more time setting up than cutting, because the batch ends before the setup pays off. Setup is not overhead in a job shop. It is the main event.
That single fact is why most scheduling advice does not fit you. It was written for stable demand and predictable routings, and you have neither.
Why static schedules keep breaking
Three things make an HMLV schedule come apart, and they compound.
First, setup dominates and setup is variable. A part you have run before sets up fast. A new one fights you. If your schedule treats every changeover as the same block of time, it is wrong before the day starts.
Second, the mix changes under you. A hot order lands, a customer moves a date, an inspection kicks a job back. In a high-volume plant those are exceptions. In a job shop they are Tuesday. Every one of them re-sequences the floor, and a plan that assumed a fixed order cannot absorb it.
Third, the estimates are optimistic. Most shops schedule from cycle time, the cutting number, which ignores load, deburr, inspection, and the walk to the tool crib. On short HMLV runs that invisible work is a big share of the job, so the schedule promises capacity that does not exist. We pulled that apart in why jobs run late even when the machines are fine, and it hits HMLV shops hardest because the runs are short and the overhead is proportionally huge.
How to run HMLV well without an APS system
Advanced Planning and Scheduling software exists to solve exactly this, and for a large plant it is worth it. For a shop with 3 to 30 machines, the price, the setup, and the weight of features you will never touch make it the wrong buy. The good news is that most of the benefit comes from three habits, not a solver.
Group jobs by setup, not by due date
The biggest lever in a job shop is doing fewer changeovers. Look for jobs that share a fixture, a material, or a tooling setup, and run them back to back so you change over once and cut several times. Sequencing by setup family instead of strictly by due date can free up real hours, which is capacity you already own and were throwing away one changeover at a time. This is the same hidden capacity that lives in setup and changeover, and in HMLV it is the single richest vein.
The catch is you cannot group what you cannot see. You need every waiting job visible in one place before you can spot the ones that belong together.
Schedule from process time, and buffer where the variance is
Time your common part families honestly and schedule from door-to-door process time, not the cutting estimate. Then, because you know your real variance, you can buffer smartly: almost nothing on a part you have run a hundred times, more on the new and fiddly work. That beats padding every job by a flat amount, which is how job shops end up quoting lead times too long to win the order. Honest numbers are also what make real capacity planning possible, because you cannot plan capacity you have never measured.
Keep one shared schedule that everyone can change
HMLV changes constantly, so your schedule has to be a living thing the whole team can see and update, not a document the owner redraws every morning. When a job slips, everyone downstream should see it move without a phone call. When a rush order lands, it goes on the same board everyone is already looking at. A shared, live schedule is what lets a small shop re-sequence on the fly without losing track of anything, which is the real skill HMLV demands.
Variety is your edge, not your problem
It is easy to look at a high-volume plant and envy the calm. Do not. That variety is why your customers come to you instead of overseas: you take the small runs, the odd parts, the rush jobs nobody else wants. The mix is the business.
The goal is not to make an HMLV shop run like a production line. It is to build a schedule that expects the chaos, groups the work to kill setup, tells the truth about time, and stays visible to everyone when the day inevitably rearranges itself. Do that and you can take on more variety, not less, without the Wednesday scramble. For a deeper walk through the mechanics, our complete guide to job shop scheduling picks up where this leaves off.