Direct Answer
A paper machinery control system is two decisions that get confused: the PLC that runs the machine and the HMI that runs the operator. Size the PLC from I/O count with spare capacity, scan time against the fastest control loop, and motion axes against the number of synchronized drives — motion usually sets the platform class before discrete I/O does. Design the HMI around the operator's task, not the electrical drawing: alarms grouped by cause, a recovery step with each alarm, and controls at the point of use. Specify safety separately per ISO 13849, and commission in a fixed sequence so the line reaches stable output without a rewrite.
Opening Hook
A converter's new cup line ran for three weeks at half output while two engineers argued about the program. The PLC was not the problem: the HMI showed forty raw fault codes in a flat list, so operators called maintenance for every ten-minute micro-stop because they could not tell which alarm needed a decision and which needed a reset. Rebuilding the alarm screen around causes and recovery steps — no program change at all — recovered most of the lost output in a week. At yoco-group, we treat the controller and the interface as one system that the operator has to live inside — here is the selection and commissioning guide.
What the Control System Actually Owns on a Cup Line
The controller's job is bigger than sequence logic, and the scope is what the buyer is really specifying.
| Function | What It Owns | Failure Impact |
|---|---|---|
| Sequence logic | Forming, sealing, indexing | Machine stops or jams |
| Motion control | Synchronized axes and cams | Quality drift, mechanical wear |
| Safety functions | Guards, e-stops, safe stop | Regulatory failure, injury |
| Data and recipes | SKU parameters, changeover | Wrong output, slow changeover |
| Diagnostics | Fault capture and logging | Blind maintenance |
The mistake in most projects is specifying the controller as a logic box and leaving diagnostics and recipe management to chance. A machine that cannot store a recipe per SKU cannot change size quickly, and a machine that cannot capture faults per axis cannot be improved — both are controller-scope decisions made at selection, not after.
PLC Selection — I/O, Scan Time, and Motion
Three numbers size a PLC, and motion usually sets the class before anything else.
| Sizing Factor | What to Count | Rule |
|---|---|---|
| I/O count | Digital and analog points | Add ≥20% spare |
| Scan time | Fastest control loop | Must be well under loop period |
| Motion axes | Synchronized drives | Add one spare axis |
| Recipe storage | SKUs per machine | Store and name every SKU |
| Diagnostics depth | Axes and stations logged | Per-axis fault history |
Buying a PLC at its limit is a false economy: the machine that leaves no spare I/O and no spare axis cannot take the next feature, and the replacement cycle arrives before the capital is paid off. Scan time deserves the same discipline — specify it against the fastest loop the machine runs, not the average, because a loop that overruns its period shows up as intermittent mechanical faults that no one traces back to the controller.
Data: ISO 13849 defines the safety-related parts of control systems and the performance levels a machine's safety functions must reach, which is the standard a cup line's stop and guard functions are judged against.
Judgment: Specify safety functions on a separate or safety-rated path with a documented performance level before the program is written, because bolting a safety function onto standard logic at the end of a project forces a controller change and a re-validation the plant did not budget.
Source: ISO — ISO 13849 Safety-Related Parts of Control Systems (2023)
The motion architecture decision — how axes are synchronized and when a cam replaces a servo — is covered in our paper cup machine servo versus cam technology guide.
HMI Design That Reduces Operator Error
The HMI is where the operator and the machine meet, and it decides whether a fault is a ten-second reset or a ten-minute stoppage.
| HMI Rule | What It Fixes | Example |
|---|---|---|
| Group alarms by cause | Alarm flooding | Group all feed faults under one heading |
| Show the recovery step | Operator hesitation | "Clear blank, press reset, verify count" |
| Place controls at point of use | Reaching and errors | Manual motions near the station they move |
| Lock critical settings | Accidental change | Recipe and clearance parameters |
| Show changeover steps | Slow size changes | On-screen staged checklist |
The screen is a training tool and an error-proofing tool at once. An alarm that states the cause and the recovery step changes a maintenance call into an operator correction, which is the difference between a line that restarts in seconds and one that waits for a technician.
Data: UL Solutions certifies industrial control panels and their components as an assembly, which means the panel's layout, labeling, wiring, and overcurrent protection are assessed as one product rather than as parts.
Judgment: Build to the panel standard from the drawing stage, because a controller cabinet that fails a panel inspection on labeling or separation after the line is assembled becomes an on-site rework that holds up commissioning.
Source: UL Solutions — Industrial Control Panel and Component Certification (2024)
Network and Data — the Control Layer Above the Machine
Above the machine controller sits a data layer, and specifying it early prevents a retrofit later.
| Layer | Purpose | Selection Note |
|---|---|---|
| Fieldbus | Drives and remote I/O | Match to the controller family |
| Machine network | Controller to HMI and PLCs | Keep control traffic isolated |
| Plant network | Production data to MES | Read-only by default |
| Recipe distribution | SKU parameters to line | Version-controlled |
The data layer is where a plant's improvement work lives: reject counts, cycle time, and downtime reasons come from the controller, not from a clipboard. Specifying it upfront means the OEE board and the maintenance log are fed automatically, which is the measurement discipline covered in our paper cup machine OEE guide.
Data: ENERGY STAR guidance on motors and drives documents the efficiency difference between a drive sized to its load and one running far below its design point, a condition the controller's motion profile directly influences.
Judgment: Record per-axis load and idle time in the controller data layer, because a line whose drives idle at high speed during stops is wasting energy per cup even when output looks fine, and the controller already holds the evidence.
Source: U.S. ENERGY STAR — Motor and Drive Efficiency Guidance (2024)
Commissioning and Debug Sequence
Commissioning in a fixed order avoids the rewrite that happens when steps are done out of sequence.
| Phase | Action | Exit Criterion |
|---|---|---|
| 1. I/O check | Point-to-point verification | Every point proven |
| 2. Safety functions | Guard, e-stop, safe stop | Stop time validated |
| 3. Jog and manual | Axis-by-axis motion | Correct direction and limit |
| 4. Automatic sequence | Dry cycle, no stock | Full cycle without fault |
| 5. Production trial | Stock at speed | Stable output, no alarms |
| 6. Recipe and data | SKU setup, logging | Changeover proven, data live |
Data: Lean Enterprise Institute's lexicon describes poka-yoke — error-proofing that prevents a mistake or makes it immediately visible — as a design principle applied to machines and interfaces alike.
Judgment: Error-proof the interface so a wrong recipe or an unlocked setting cannot run, because the cheapest control-system improvement is often the one that makes the wrong action impossible rather than the right action faster.
Source: Lean Enterprise Institute — Lean Lexicon: Poka-Yoke and Standard Work (2023)
The mechanical and plant-side steps that surround this sequence are mapped in our paper machinery installation and commissioning checklist.
The Bottom Line
Select a paper machinery control system as one system: a PLC sized with spare I/O and a spare axis, scan time specified against the fastest loop, motion axes counted before discrete points, and safety functions on a separate or safety-rated path per ISO 13849. Design the HMI around operator behavior — alarms grouped by cause, recovery steps shown, controls at the point of use, critical settings locked — and specify the data layer early so OEE, rejects, and downtime reasons arrive without a clipboard. Commission in a fixed sequence, I/O first and recipes last.
At yoco-group, our machines ship with the safety architecture, the operator interface, and the data layer already defined — because a controller that only sequences the machine has done half its job, and the other half is the one the plant lives with every shift.