Asset Health
Asset Health
Asset Health is the Shop Floor module where maintenance and reliability teams see every alarm and fault, follow each line station by station, check the health of each machine, and turn recurring problems into Plex maintenance work requests.
Permissions: Asset Health needs the Asset Health license on top of your Shop Floor license. Without it, the module's menu entries and boards do not appear. Asset Health has its own roles. DataMagik - Builder and DataMagik - Viewer give you nothing here. DataMagik - Shop Floor Viewer opens every screen and can acknowledge alarms. DataMagik - Maintenance can do all of that, and can also work insights: change their state, add notes, create insights, raise a Plex work request, act on suggested work requests and mark a PM done. You do not need Shop Floor Viewer as well. Only DataMagik - Shop Floor Admin sees Settings and can change layouts, photos and health profiles. If you hold Maintenance but not Shop Floor Admin, you start on the Asset Health Portal when you sign in. Your company admin assigns these roles in the user manager.
Where to find it
- Shop Floor menu: the Shop Floor dropdown has an Asset Health group with Portal, Lines, Assets, Insights, Analytics, Tag Explorer and, for a Shop Floor Admin, Asset Health settings.
- Shop Floor home: Portal, Lines, Assets, Insights and Analytics appear as boards next to your other Shop Floor boards. Each one opens the module for the location you chose. See Shop Floor boards.
- Inside the module (
/shop-floor/asset-health): a breadcrumb back to Shop Floor, a Location picker when you have several locations, and a tab strip with Portal, Lines, Assets, Insights, Analytics and Tag Explorer, then any linked apps, then Settings (Shop Floor Admin only).
Health statuses
Every machine has one of five statuses. Each status is shown as a coloured dot and a word, never colour alone.
- Operating (green). The machine reports and nothing is wrong. Line and station views say Running.
- Warning (amber). An open warning-level incident is on the machine.
- Alarming (red). An open critical incident is on the machine. Line and station views say Faulted.
- Stopped (blue). The machine's run state says it is not producing (down, idle, fault, starved, blocked, setup, no batch or no job, or a break), and it has no critical incident.
- Offline (grey, with a dashed ring). No fresh signal has arrived, the signal dropped out, or the sensor is frozen.
How the status is worked out
The status is worked out each time a screen asks, never stored. For an asset, the first rule that applies wins:
- Offline: silent longer than its Offline after setting (5 minutes by default), its signal dropped out, or its sensor is frozen, which means it keeps sending the same value on every channel. Stale data cannot be trusted, so this outranks an alarm.
- Alarming: an open critical incident.
- Stopped: its run state, or its linked workcenter's status, says it is down, idle, fault, starved, blocked, setup, no batch or no job, or on a break.
- Warning: an open warning incident. An info incident changes nothing.
- Operating: none of the above.
A workcenter's status comes from its state or status value: setup, idle and down read as Stopped, and no value or a stale value reads as Offline. Open incidents then apply as above. Lines, areas and assets with components show the worst status below them, with Alarming ranked above Offline, so a parent points to where to look first.
Equipment reports its run state through the signal machine_state. Bind its run-state tag or historian series under that name: role names such as state are reserved and cannot be used for an equipment signal.
Incidents come from conditions. See Conditions and escalation, and set asset thresholds in Asset Health setup.
Portal: the alarm and fault portal
The Portal is the plant-wide view.
- KPI strip: Active faults, Warnings, Lines running (for example 3 / 4) and Machine downtime today. Machine downtime is summed over every machine, so it can exceed the hours in the day; the tile says how many machines it adds up. At a desk, the strip also shows MTTR today and Top fault (30 days).
- Lines: one card per line, with its status, faults, warnings, downtime today and availability. Each card also says why: its machines counted by run state, with the downtime reason, for example "21 starved · 2 down (Preventive maintenance)". Select a card to choose that line, or Open line detail → for the Lines screen.
- Active alarms: every open alarm at the location, with how long each has been active. When a line is chosen, only that line's alarms are listed, with a Show all lines link.
- Alarm detail: follows the alarm list, so it always describes an alarm the list shows. It gives the selected alarm's Possible causes and Recommended checks, an Acknowledge button, and, when the alarm is on equipment, Add note and Create Plex work order on the insight it belongs to. Alarms on a workcenter have no insight, and the panel says so.
- Top faults: a pareto of downtime by status reason over the last 30 days, longest first, with the cumulative share. Downtime with no reason is reported separately.
- Downtime by line: today's downtime and number of stops for each line. Select a bar to choose that line.
- Availability trend: each line's daily availability over the last 7 days (planned time minus downtime, over planned time, from Plex status events, or for lines of PLC equipment from their run-state history). Select a legend entry to choose that line. Show as a table gives the figures as a table.
Choosing a line
Choose a line anywhere on a screen and every screen narrows to it. You can choose a line from a line card, a Downtime by line bar, an Availability trend legend entry, the line schematic or the asset tree.
- Active alarms lists only that line's alarms, on every screen, with a Show all lines link. The alarm detail follows the narrowed list.
- Downtime by line and the Availability trend mark the chosen line and fade the others. Select the chosen bar or legend entry again to clear it.
- The Assets carousel and the Insights list narrow to that line's equipment, with a link such as Only Press Line 1 · show all to see everything again.
Equipment downtime and availability
A line of PLC equipment has no Plex status events, so for those lines Downtime by line, the line cards, the KPI strip's Machine downtime and the Availability trend are worked out from each cell's run-state history in your historian:
- Running counts as operating.
- Break, Lunch and Off are outside planned time, so they count neither for nor against availability.
- Any other state, such as down, starved, blocked, setup or no batch, is downtime. Back-to-back downtime states count as one stop.
This needs a historian series mapped to each cell's machine_state signal. A cell without one adds nothing to its line's downtime or availability. See SQL historian.
Lines
The Lines screen follows one line, picked from a list.
- Line schematic: the line's stations in material-flow order, left to right, each with its photo or icon, its status and its live Key signal, named in words, for example "Motor rpm 1482". A stale value says Stale; a station with nothing reporting says No signal.
- Station detail (desk): select a station to see its open alarm and key signal.
- Today by station (desk): a timeline of today for each station. A workcenter station shows its Plex status events. An equipment station shows its run state, read from your historian through the
machine_statesignal. Break and Lunch are grey and count as stopped. - The Lines screen also shows the KPI strip, the line's own active alarms with their detail, and the availability trend.
Assets
- Asset health summary: counts by status. Select a tile to filter the tree and lists to that status.
- Asset tree: location, areas and lines, with workcenters and assets together under each line, each parent showing its roll-up status. Use Search asset or location to find one.
- Asset detail: photos, health, Health profile, a Sensors table (signal, source, value, as of, threshold) and Open insights.
- Trending toward a limit and Upcoming PM lists on the Assets screen show the assets heading for a limit and the PM that is due soon or overdue.
- Needs attention (wall layout): a carousel of alarming and warning assets, worst first, showing how long each has been in that state, for example "Alarming for 3d 11h".
What the asset detail shows
- Health score: a score out of 100 with its reason and its change over 7 days.
- Signals against normal: each signal compared with its learned running normal. A stopped motor is not compared.
- Likely cause: a line such as "Likely: bearing wear · because ...", decided by a fixed set of rules. Other causes include unbalance or misalignment, looseness or impacts, and a load or process change. If your company has AI Agents, select Explain to have the assistant put it in plain words. The cause itself does not change.
- ISO 10816 zones: RMS velocity shows its zone (Good, Satisfactory, Unsatisfactory or Unacceptable), judged only while the motor runs.
- Trending toward a limit and Reaching Unsatisfactory soon: forecasts of when a signal will reach its limit at the current rate.
- Spectrum: vibration bands, including bearing bands, against their learned normals. A Shop Floor Admin sets the running speed and the bearing.
- Energy: the asset's energy use.
- Run-hour maintenance: the next and any overdue PM, from the Plex PM schedule or from a DataMagik run-hour interval. A Shop Floor Admin can set a DataMagik interval for equipment Plex does not schedule. A user with Maintenance can select Mark PM done. On a label printer's equipment this section is Printer maintenance, counted in labels or km of printhead instead of run hours. See Printers as Shop Floor equipment.
Insights
An insight is one problem on one asset. When a condition opens an incident on an asset, it joins that asset's open insight or opens a new one named after the condition, so a bearing that alarms four times this week is one insight with four alarms.
- Filter the list by Open, Pending, In inspection, Closed or All, and select an insight.
- Move it through Pending → In inspection → Closed. You can also reopen it.
- Read the Linked alarms with their causes and checks, and add notes under Notes and history.
- Select Create Plex work order…. The first step is always a dry run: nothing is sent to Plex. The page shows the data source, the priority, every field that would be written, and anything that would stop the send.
- Select Confirm and create in Plex to raise the work request. A pending insight moves to In inspection, and the work request number is shown on the insight.
An insight gets one work request; a second confirm is refused. Priority comes from the asset's criticality and equipment from its Plex equipment, and the dry run names anything missing.
A user with Maintenance can also open an insight by hand with + New insight. An insight shows This has happened before when earlier insights were opened on the same equipment.
After a work request is raised, the insight reads it back from Plex every 30 minutes. It shows who completed it and the repair time. It then waits for 48 running hours after completion and says whether the fix held: Fixed, Partly fixed, Not fixed or Not judged.
Suggested work requests
Above the Insights list, and on the asset card, Suggested work requests lists drafts Asset Health has raised for you. Each one says why: PM due, Fix didn't hold, Unacceptable vibration, Reaching Unsatisfactory, Health dropping or Printer maintenance due. Nothing goes to Plex until someone confirms it.
- Select Preview work request…. This is a dry run, and nothing is sent to Plex.
- Select Confirm and create in Plex to raise it.
- Or select Dismiss…, say why, and select Dismiss suggestion.
Acting on a suggestion needs Maintenance. A Shop Floor Admin sets how far ahead a motor forecast to reach ISO Unsatisfactory is suggested, as a Lead time (days, 1–30) under Insights & work orders in settings.
Analytics
- Signal history: pick an asset to chart each signal mapped from your historian, over 8h, 24h, 7d or 30d. The line is the average, the band is the minimum to maximum, and threshold lines come from the asset's conditions. History is always read from your own historian. See SQL historian.
- Reliability for the same window: Failures, MTBF, MTTR, Availability and Time in alarm. A failure is a critical incident that opened in the window, and downtime is the time critical incidents cover. MTBF is uptime divided by failures; MTTR is the average time to resolve a failure. Time in alarm counts incidents of every severity. Incidents raised by a sensor fault are not counted, and show as Sensor fault (not counted).
- Analytics also shows the KPI strip, Top faults, Downtime by line and the Availability trend.
Management
At a desk, Analytics also has a Management section. Choose a Window of the last 7, 30 or 90 days. The section is read-only, and its figures are kept for five minutes.
- Fleet health by line: each line's health score now against 7 days ago, an 8-week trend, and the motors forecast to reach ISO Unsatisfactory within 7 or 30 days.
- PM compliance: PM done on time, late or still overdue.
- Repair effectiveness: repairs that held and repeat failures.
- True availability by line: availability as Plex recorded it against true availability, which also counts the stops the motors show that Plex logged as running.
- Fleet comparison (Beta): equipment of the same type side by side.
- Energy per part by line: kWh per good part against normal.
- Part profiles (Beta): scrap, rate and motor normals for each part.
Alarm quality
Below Management, Alarm quality covers the last 90 days. For each condition it shows how often it Fired, how many firings Led to a work request, and how many of those Repairs held. Missed faults lists work requests that no alarm came before.
Where the history supports it, a condition gets a suggested new limit with its evidence. A Shop Floor Admin can select Preview a change and then Apply. Nothing changes until someone applies it.
Tag Explorer
Tag Explorer puts historian tags side by side, so you can see how they move together.
- Pick up to eight tags. Use Search tags, equipment or lines, or browse by line, equipment and signal.
- Choose a Time range: 8 h, 24 h, 7 days, 30 days or Custom. All tags share one time axis. Tick Normalized 0–100% to show each tag against its own range.
- Read Relationships: the correlation of each pair, and the time lag where they line up best. Strong pairs are described in a sentence. Select a cell to see that pair as a scatter plot.
Correlation is not cause. Two tags that move together may both follow something else.
Reading charts
Hover over any chart to read the value at that moment; time charts show a guide line with each series' value. With the keyboard, select the chart and use the Left and Right arrow keys to step along it (Home, End and Esc also work).
Freshness
At a desk, each panel shows Live · updated Ns ago. A wall board shows one freshness label for the whole board instead. Live panels refresh every 30 seconds at a desk and every 15 seconds on a wall. Top faults and the Availability trend refresh every 5 minutes. If a panel stops refreshing, it turns Stale and dims its figures, so old data never looks healthy.
Desk mode and wall mode
Every screen has a Desk mode and a Wall mode button. Wall mode is dark, with much larger type and fewer panels, and nothing in it depends on hovering. A Shop Floor Admin can change each mode's layout with Customize layout. See Asset Health setup.
On displays
Portal, Lines, Assets, Insights and Analytics can each be a rotation item on a paired display: Asset Health · Alarm & Fault Portal, · Lines, · Assets, · Insights and · Analytics. Each item can be pre-filtered to a Location, a Line and/or an Asset, chosen from pickers. A display draws the screen in wall mode, with no module menu, and shows only its own location's data. A display with no location of its own is held to the location its rotation item names. See Shop Floor displays.