Your HMI Says Step 40 — It Should Say What Step 40 Is Waiting For

The call usually starts the same way: "the machine isn't moving." You open the screen and it says Auto Running, or Step 40. That is the whole message.

The operator already knows it stopped. What they need is what it is waiting for — a valve feedback, a temperature permissive, a downstream ready signal, or an acknowledgement button they are supposed to press themselves. Sequence display design mostly collapses into that one question. State, step, and hold reason are three different pieces of information, and where the hold reason comes from while the sequence is holding is the part that decides whether the screen is useful.

The minimum a sequence screen owes the operator

A useful sequence display does not try to show the whole PLC program. It shows the current operating state, the current step, the expected next action, and the reason the sequence is holding. Everything beyond that belongs on a maintenance screen.

• unit or equipment name;
• mode: manual, auto, semi-auto, maintenance, simulation if allowed;
• sequence state: idle, starting, running, holding, completing, aborting, faulted;
• current step number and step name;
• active hold reason or first blocking condition;
• elapsed time in the current step;
• safe recovery action.

If the screen only exposes a raw integer step, the operator still needs the PLC printout or a controls engineer. That is not a good operating interface.

Where that list lives is also settled. In the display hierarchy of ANSI/ISA-101.01-2015, sequence detail sits at Level 3 — the unit detail display. Push hold reason text up to the Level 1 overview and the overview stops being an overview. Bury it in a Level 4 diagnostic display and the operator will never see it.

Keep step numbers and names together

Step numbers are useful for engineers because they match PLC logic, commissioning notes, and alarm records. Step names are useful for operators because they describe the process.

Show both:

Step 40 - Wait for Fill Valve Open Feedback

Avoid showing only:

Step = 40

Also avoid hiding the number completely. During troubleshooting, a controls engineer may ask, "What step is it stuck on?" The answer should be visible without opening a diagnostic tool.

Keep step names stable. Renaming Heat Soak to Temperature Stabilization looks harmless, but it breaks procedures, training notes, and shift handover language. If names must change, update alarm text, SOP references, and historian event labels in the same change. Fix one place only and in six months nobody knows the two names are the same step.

Separate state, step, and hold reason

Three different things are often mixed into one display field:

Item	Example	Purpose
State	Holding	What the sequence controller is doing now
Step	Step 40 - Fill	Where the sequence is in the recipe or cycle
Hold reason	FV-102 open feedback not made	Why progress is blocked

Mixing them produces vague messages like Sequence Fault or Waiting. A better display is:

State: Holding
Step: 40 - Fill Tank A
Hold: FV-102 open feedback not made after 8.0 s
Next: confirm air supply and valve limit switch, then reset hold

If the state names are homegrown, map them once against the machine state model in ISA-TR88.00.02, the technical report most people just call PackML. There, HOLDING and HELD are different states: HOLDING is the transition into hold, HELD is sitting there. Collapse both into one word and the screen can no longer distinguish "stopping right now" from "already stopped" — which is exactly where the operator decides whether to keep waiting or to intervene.

The HMI does not need to replace alarms. It should make the active cause visible near the sequence controls so the operator does not hunt across four screens.

Use first blocking condition carefully

A sequence may have many permissives. Showing all of them at once floods the operator with normal inactive conditions. Showing only a generic hold hides the useful detail.

A practical approach is to show:

1. The first blocking condition used by the PLC for sequence progress.
2. A compact list of other active blockers if more than one exists.
3. A detail button that opens the full permissive or interlock view.

The "first" condition should come from deterministic PLC logic, not from whichever SCADA tag updated first. If SCADA derives the first blocker from tag arrival order, the answer changes with scan timing and network delay — and a screen that names a different cause each time the same fault occurs is worse than no screen.

For packaged equipment, ask the vendor what priority order the controller uses. If none exists, agree on one during commissioning and document it.

Show timers as engineering information, not decoration

Timers help an operator decide whether the sequence is progressing normally. Useful timer displays include:

• elapsed time in current step;
• expected or configured timeout;
• remaining dwell time for soak, purge, drain, or cool-down steps;
• timeout alarm threshold;
• last successful cycle time for comparison when available.

Example:

Step elapsed: 00:01:18
Fill timeout: 00:02:00
Expected fill time: 00:00:55 to 00:01:20

A timer should have units and meaning. A bare number like 78 is not enough. If a timer resets during step transitions, the screen should not freeze the old value without marking it as previous-step data.

The last item on that list — the previous good cycle time — turns out to be the one operators use most. "Forty seconds left before timeout" moves nobody. "Already twice as long as normal" moves everybody.

Common failure modes

The HMI shows the next step too early

Some PLC programs load the next step number before the action is complete. The operator sees Step 50 - Agitate while the equipment is still filling. This creates confusion during startup and fault recovery.

Field check: compare the displayed step with real outputs and sequence state during a dry cycle. If the step number represents "commanded next step" instead of "active step," label it that way or expose a separate active step.

Hold reason clears before the operator reads it

A feedback flickers, the hold reason appears for one scan, and then disappears. The sequence may still fail later, but the useful clue is gone.

Latch the last hold reason with its timestamp until the next sequence start, reset, or operator clear — and state on the screen and in the procedure which of those three clears it. Latch it unconditionally and you get the opposite problem: last week's hold reason still sitting on the display.

Manual recovery is not tied to sequence context

After a fault, an operator may need to jog a valve, drain a vessel, or return an actuator home. If the recovery controls are far from the sequence step display, the operator can recover the wrong unit or miss a required permissive.

Recovery faceplates should carry the same unit name, current step, hold reason, and command target as the main sequence screen. On a line of identical units, that is not display polish — it is a safeguard.

Simulation and bypass states look like normal ready states

Commissioning bypasses and simulation bits are sometimes hidden because engineers do not want to clutter the operator screen. That becomes dangerous when a bypass remains active after testing.

Field check: if a bypass, force, simulation, or maintenance permissive affects sequence progress, show it next to the hold reason and include it in shift handover.

Historian and event logging

Sequence troubleshooting improves when the HMI display matches stored events. Log at least:

• sequence start, complete, abort, and reset;
• step changes with step number and name;
• hold reason changes;
• timeout alarms;
• operator recovery commands;
• bypass enable and disable events.

Use source timestamps from the controller when available. For fast step changes, a historian deadband or slow scan class hides intermediate steps. Treat step number, state, and hold reason as event-like values, not slow analog trends.

Commissioning checks

Before handover, run one normal cycle and at least two controlled abnormal cases:

1. Block a feedback and confirm the hold reason matches the real blocker.
2. Trigger a timeout and confirm alarm text, sequence display, and event log agree.
3. Move through steps quickly and confirm no intermediate step is lost if it matters for diagnosis.
4. Switch from auto to manual and back only through the approved recovery path.
5. Verify the same step name appears in HMI, alarm text, historian events, and procedures.
6. Confirm Korean, English, or multilingual text does not truncate the equipment name or hold reason.

One more thing to measure before you finalise any of the wording: hold reason text hits the PLC string length long before it hits the screen width. The default STRING on Logix-family controllers holds 82 characters. FV-102 open feedback not made after 8.0 s fits, but add a unit prefix and a second language and it will not — and the overflow disappears quietly.