In the past two months working as a production controller in the metal furniture industry, I discovered an interesting phenomenon: although each process has its own sequence, the time perspective often differs completely because of the question “when does the next station need it?”
Take an order for example. From front-end processing, tube bending, and stamping to the later welding stage, each process has its own production time and corresponding BOM relationships. In theory, the previous process should prepare in advance according to the overall schedule so the next process can continue smoothly. But in actual communication with the shop floor, I often hear the same line: “Just tell me when the next station needs it.”
At first, I wasn’t sure how to respond, because from the shop floor’s perspective the statement makes perfect sense. If the next process hasn’t called for material yet, why start early? As long as the parts are ready before the other side needs them, isn’t that enough?
Yet the real difficulty in production control scheduling is that we cannot look only at “when the next station needs it.” We usually see the entire order book over a period of time. Sometimes several orders share the same due date and all appear to ship at the end of the month, so the shop floor easily thinks, “Since everything is due at the end of the month, we can just work through them in order.” But if everything really waits until the same moment to be scheduled backward, everything ends up piling up at the final process at once.
The earlier processes may feel they only need to finish before the last station requires the parts, yet the last station cannot digest that volume in a short time. The result is that everyone earlier says “we’re fine,” while the final station suddenly has to work overtime and gets chased every day asking why the parts aren’t out yet.
This is also what makes production control interesting. The shop floor sees “when I need to hand off to the next station,” while production control sees “if we don’t do it now, will we fall behind later?” So when I bring the schedule to discuss with the shop floor, they sometimes feel I scheduled things too early and that it should be enough just to tell them when the next station needs the parts. But the schedule isn’t meant to keep everyone busy today; it’s meant to clear potential bottlenecks before they happen.
It’s like a highway: if everyone knows a large number of cars will enter at the same time tomorrow afternoon, we can figure out how to divert traffic now instead of waiting until all the cars are stuck at the interchange before asking why everything is jammed.
Of course, this doesn’t mean the shop floor must follow the schedule exactly as written. Actual production always has variables—machine conditions, manpower, materials, or other urgent orders on site—that may require adjustments. A truly good schedule is never something a production controller finishes alone at a computer and then hands to the floor with orders to follow.
Instead, production control first lays out future demand and risks, then discusses with the shop floor: “Can you meet this timing? Which part looks difficult? If we move this order earlier and that one later, would it flow better?” Only after finding a way both sides can accept does the schedule become workable.
Because production control and the production line are not on opposite sides—production control wants orders delivered on time, and the line doesn’t want to be chased for last-minute parts every day; production control doesn’t want the final station to get congested, and the line doesn’t want work that could have been spread out to suddenly pile up at one moment.
So I increasingly feel that the value of scheduling is not “predicting what will happen,” but leaving a little buffer for the future in advance. The shop floor only sees the next station; production control has to see the entire line. A schedule that actually works isn’t about how pretty it looks on paper, but about production control and the shop floor discussing it together so every process can connect smoothly.
This is probably also why, the longer I do this job, the more interesting it becomes: what we schedule is never just a chart; we are arranging an entire production chain so it doesn’t crash into itself.