Every PLC in a steel plant keeps a running record of its machine: cycles completed, hours run, fault bits and the state of the line when they went high. Planners rarely see any of it and work from a calendar and a handwritten log. Moving PLC data into the maintenance workflow closes that gap, but only when each tag is chosen, cleaned and tied to an asset. This guide follows the path from controller register to closed work order. To prepare the maintenance side, create your Oxmaint account and build the asset register first.
Integration · PLC Data · Steel Plant CMMS Workflow
Steel Plant PLC Data to CMMS Workflow for Maintenance and Production Data
Turn fault bits, counters and run-hour registers into work orders, usage-based PMs and production-aware maintenance, without rebuilding the control system.
PLC register
Fault bit, counter or analog value
then
Gateway or server
Protocol bridge, buffering, quality flag
then
Rule engine
Dwell time, state check, deduplication
then
Work order
Asset, priority, evidence, owner
The Problem: Controllers Know More Than Planners Do
Without PLC data in the workflow
- PMs fire on the calendar, not on how hard the machine worked
- Breakdown requests arrive by phone with a vague description
- Downtime reasons are typed in later from memory
- Technicians arrive without the fault history
- Production and maintenance argue about what actually stopped the line
With PLC data in the workflow
- PMs trigger from counters and run hours on the assets that earned them
- Faults raise a work order carrying the tag, time and operating state
- Stop events are tied to an asset and a work order record
- Technicians see the alarm sequence before they open the cabinet
- Both teams read the same event record
Five Kinds of PLC Data and What Each Is Good For
Treat PLC data as five families. Each one drives a different maintenance behavior, and mixing them is what makes early integrations noisy.
| Data family | Typical steel plant example | Workflow it drives | Main risk |
|---|---|---|---|
| Fault and trip bits | Drive fault on a mill motor, overload trip on a pump starter | Corrective work order | Operator resets hide repeat faults |
| Counters | Cycles on a ladle slide gate, strokes on a hydraulic cylinder, coils processed | Usage-based preventive maintenance | Counter rollover or reset after a program download |
| Run-hour accumulators | Fan, pump, conveyor and compressor hours | Meter-based PM and lubrication tasks | Hours counted while the motor is energized but idle |
| Analog process values | Bearing temperature, hydraulic pressure, motor current, cooling water flow | Condition-based inspection or work order | Sensor drift mistaken for machine wear |
| State and mode bits | Running, idle, changeover, maintenance mode, E-stop | Context for every other signal | Missing context creates false work orders |
A Reference Workflow in Seven Moves
Automation engineer
Exposes agreed tags through a server or gateway and documents data type, units and scaling for each one.
Reliability engineer
Decides which tags matter for which failure mode, and defines thresholds, dwell times and priorities.
Maintenance planner
Owns the work order templates, parts lists and job plans that a triggered event will use.
Maintenance system
Receives the qualified event, matches it to the asset, raises or updates the work order and notifies the right crew.
Technician
Opens the job on a mobile device, reads the attached trend, completes the checklist and records the failure cause.
Production supervisor
Sees which assets are down or at risk and agrees access windows for planned jobs.
Reliability review
Compares triggers with outcomes each month and retunes thresholds that raise too many or too few jobs.
Data Quality Gates Before a Tag Reaches a Work Order
Most integration problems are data quality problems. Put these gates between the PLC and the work order queue.
Gate 1
Quality
Ignore values flagged bad, stale or uncertain. A lost network link must not read as a fault.
Gate 2
Time
Keep the source timestamp and check clock sync, so alarm order across controllers can be trusted.
Gate 3
State
Suppress triggers during planned stops, changeovers and maintenance mode.
Gate 4
Dwell
Require a condition to persist, or repeat a set number of times, before it counts.
Gate 5
Duplicate
If a work order is already open on the asset for the same failure mode, add the event to it.
Start With One Line, Ten Tags and a Review Date
Prove the workflow on a single production line before you copy it. Set up the assets, templates and PM schedules in Oxmaint and test the triggers in review mode.
Trigger Rules: Worked Examples in Plain Language
These are illustrative patterns. Thresholds and intervals must come from your own equipment manuals, history and reliability review.
If a main drive fault bit is active while the line state is Running
Create a corrective work order on the drive, high priority, attach the last stretch of current and speed trend, and notify the on-shift electrician.
If a pump overload trip occurs more than a set number of times in a rolling week
Raise an inspection work order on the pump and motor, and link the repeat count so the planner sees the pattern, not one event.
If an actuator cycle counter reaches its service interval
Generate the usage-based PM with its job plan and parts list, then reset the next-due reference from the actual counter reading.
If bearing temperature stays above the inspection limit beyond the dwell time
Request an inspection round, ask the technician to confirm with a handheld reading, and escalate only if the reading agrees.
If run hours on a fan or compressor reach a lubrication interval
Create the lubrication task for that asset, so greasing follows use instead of the calendar.
Connecting Production Context to Maintenance Records
Maintenance data becomes more valuable when it carries the production situation around it. A fault during a grade change means something different from the same fault in steady running.
What to capture with each event
- Line state and operating mode at the time of the event
- Product, heat, coil or campaign reference where the plant records it
- Start and end of the stop, from the PLC rather than from memory
- Which asset caused the stop, and which were only affected
What this makes possible
- Usage-based PM tied to tonnage or cycles, not to elapsed days
- Downtime reports that maintenance and production both accept
- Planned work scheduled into genuine low-load windows
- Repeat-failure analysis by operating condition
Where Steel Plant PLC Integrations Go Wrong
| Mistake | What happens | Better practice |
|---|---|---|
| Mapping every available tag | Thousands of events and no planner capacity to read them | Select tags that answer a failure-mode question |
| No asset hierarchy first | Events cannot be matched to equipment, so they land in a general queue | Build the asset register and criticality ranking before mapping |
| Ignoring counter resets | PM intervals jump after a program download or controller replacement | Handle rollover and reset events explicitly in the rules |
| Treating all alarms as faults | Nuisance alarms flood the work order list | Rationalize alarms and use dwell time and state checks |
| No closed loop | Triggers are never reviewed against outcomes | Require failure codes at closure and review monthly |
| Changing PLC code without notice | Tags move or change meaning, and the integration breaks quietly | Add the tag map to the plant change process |
Plant Areas Where PLC-Driven Workflows Pay Off First
Material handling
Conveyor run hours, belt trips and pull-cord events for usage-based PM and fast response.
Rolling mill auxiliaries
Hydraulic units, lubrication systems and descaling pumps with pressure and trip history.
Continuous casting
Cycle counts on actuators and drive faults where a stop is costly and spares need planning.
Utilities
Compressors, cooling water pumps and fans with run hours, temperatures and start counts.
Cranes and handling
Hoist starts, brake events and limit faults tied to inspection and component life.
What Oxmaint Handles on the Maintenance Side
Oxmaint supplies the maintenance workflow that PLC events feed into. The connection method depends on your controllers, network design and the integration approach agreed for the site.
- Asset management Hierarchy, criticality and history for every machine that has a mapped tag.
- Work orders and PM Corrective, calendar, meter-based and condition-based jobs in one queue.
- Mobile workflows Technicians receive jobs, read evidence, complete checklists and record failure codes in the field.
- Inventory Parts attached to assets and job plans, so a triggered PM already lists what it needs.
- Reporting Backlog, PM compliance, repeat failures and downtime-related work on dashboards.
Readiness Scorecard
Answer yes or no. Three or more no answers mean the integration should wait until those basics are fixed.
Critical assets are ranked and uniquely identified
The PLC tag list is documented with units and scaling
A named owner exists for each controller's data
Planned stops can be detected from a state tag
Job plans exist for the first ten triggered tasks
Failure codes are defined for closing work orders
Network zones and access rules are agreed with plant IT
A monthly review slot is booked for trigger tuning
Choosing a Connection Method for PLC Data
There is no single best method. Match the method to the controller age, the network layout and who owns the data.
| Method | Typical use | Strength | Watch-out |
|---|---|---|---|
| OPC UA server | Controllers and SCADA packages that expose a standard server | Structured, secure and vendor-neutral | Tag names need mapping to assets |
| Vendor driver or gateway | Older controllers with proprietary or serial protocols | Reads existing registers without changing the PLC | Another device to document and maintain |
| Edge gateway with broker messaging | Remote units and several sites | Buffers data when links drop | Topic design and ownership must be agreed |
| Historian or database query | Trends and event history already stored centrally | No extra load on controllers | Only as good as what is stored and how often |
| Scheduled file export | Systems with no live interface | Simple, and a reasonable first step | Slow, so unsuited to urgent faults |
Test Cases to Run Before Real Work Orders Flow
Test: a fault bit is forced in a safe test condition
Expected result: one work order is created on the right asset, with the right priority, timestamp and attached values.
Test: the same fault repeats ten times in a short period
Expected result: the events join one open work order, with a repeat count, instead of ten separate jobs.
Test: the line is placed in maintenance mode, then a fault is forced
Expected result: no corrective work order is raised, because the state check suppresses the trigger.
Test: the network link is cut and restored
Expected result: no false faults on reconnect, and buffered events appear once and in the right order.
Test: a counter is reset as it would be after a program download
Expected result: the PM schedule handles the reset deliberately and does not jump forward or backward.
Counters, Resets and Rollover
Counters and run-hour registers are the cheapest route to usage-based PM, and also the easiest to get wrong.
- Record the data type and maximum value of each counter, so rollover can be detected rather than misread as a drop to zero.
- Treat a sudden fall in the value as an event to investigate, not as normal use.
- Store the last good reading in the maintenance record, so the next-due calculation survives a controller replacement.
- Separate energized time from running time where the PLC allows it, because a motor that is on but idle does not wear like one under load.
- Reconcile counter-based intervals with a manual reading during the first PM cycle on each asset.
Change Management When Controller Programs Change
Before a PLC change
- Check whether the change touches any mapped tag, address or data type
- Tell the maintenance system owner which assets will be affected
- Agree a window and a rollback plan for the integration
- Pause automatic triggers on the affected assets if needed
After a PLC change
- Re-test every mapped tag on the changed controller
- Confirm units, scaling and counter behavior are unchanged
- Update the tag map with the new versions and dates
- Resume triggers and watch the first day of events closely
Reporting Once the Data Is Flowing
Once events and work orders share an asset record, the reports that were hard to build become routine.
- Stops by asset and cause, with the time from fault to work order creation and from creation to completion.
- Usage-based PM compliance, showing which assets crossed their counter or run-hour limit and whether the job was done.
- Repeat fault lists, ranking assets by how many times the same tag has triggered in a rolling period.
- False trigger review, listing work orders closed with no fault found, so thresholds can be tuned.
- Planned versus unplanned work on integrated assets, tracked month by month against the baseline set before go-live.
Frequently Asked Questions
Do we need to replace old PLCs to feed a CMMS?
Usually not. A gateway or driver can read registers from older controllers and republish them. Book a demo to review your controller mix.
Which PLC data should become a work order?
Reliable fault bits and persistent condition breaches on critical assets. Counters and run hours are better used to schedule preventive maintenance.
How do we stop duplicate work orders from repeated alarms?
Match each event to the asset and failure mode, and add it to an open work order instead of creating a second one.
Can PLC data support usage-based PM?
Yes. Cycle counters and run-hour registers can set service intervals from real use. You can sign up and set up meter-based PM schedules.
Who should own the tag mapping?
Automation owns the tag definitions, reliability owns the trigger logic, and planning owns the job plans. One named person should own the map as a whole.
Let the Controller Start the Job, Let the Planner Finish It
Connect PLC events, counters and run hours to preventive maintenance, work orders and asset history in one maintenance workflow.
