Alarm
Table of Contents
- Overview
- Four Alarm Handlings — Acknowledge, Hide, Suppress, Out of Service
- Screen Structure — 5 Sub-menus
- Alarm Screen (
/alarm/index) - Alarm Search Screen (
/alarm/advsearch/index) - Alarm Analysis Screen (
/alarm/analysis/index) - Alarm Performance Screen (
/alarm/performance/index) - Advanced Alarm Screen (
/alarm/config/index) - Alarm Detail Screen
- Utilization Scenarios
- FAQ
- Related Screens
Overview
The Alarm group is an area that bundles real-time monitoring, history search, rule registration, visualization analysis, and performance evaluation of occurred alarms. The handling of alarms—Acknowledge, Hide, Suppress, and Out of Service—follows the status distinctions of ISA-18.2.
| Sub-menu | Internal URL | Purpose |
|---|---|---|
| Alarm | /alarm/index | Real-time alarm monitoring + Search period list + Statistics/Ranking |
| Alarm Search | /alarm/advsearch/index | Detailed search by period, range, importance, activity, acknowledgement, and content |
| Advanced Alarm | /alarm/config/index | Alarm setting list, EQL alarm registration/deployment, temporary hide, out of service |
| Alarm Analysis | /alarm/analysis/index | Visualization analysis such as ranking, trend, distribution, and treemap |
| Alarm Performance | /alarm/performance/index | Alarm system performance evaluation (ISA-18.2 / EEMUA 191) |
An unacknowledged alarm count is displayed as a red badge next to the alarm bell icon in the top header; clicking it navigates to the Alarm screen. If there are currently hidden occurrences, a separate «Suppressed N» line appears in the bell dropdown.
Four Alarm Handlings — Acknowledge, Hide, Suppress, Out of Service
One Acknowledgment per Alarm
Acknowledgment is a record that «someone has seen and taken responsibility for this alarm». It is not tracked per person, but as one record per occurrence. The person who acknowledges first remains the acknowledger, and subsequent acknowledgments do not change the initial record. Once one person acknowledges, it appears as acknowledged for everyone.
| When | What is recorded |
|---|---|
| Acknowledged via console | Acknowledger = Logged-in user |
| Acknowledged via field app | Acknowledger = That employee |
| Opening alarm details | Nothing is recorded — acknowledgment is an intentional act of pressing a button |
An alarm does not disappear after acknowledgment. If the value has not yet returned to normal, it remains active.
| Role | Permitted Actions |
|---|---|
| ADMIN | [Acknowledge Alarm] in Alarm Details, [Acknowledge All Alarms] in Alarm, Alarm Search, and Details |
| API | [Acknowledge] in the alarm detail recipient panel. Individual, batch, and total acknowledgment via V5 |
| MANAGER · USER | Acknowledge alarms one by one using [Acknowledge Alarm] in Alarm Details. [Acknowledge All Alarms] is not visible |
[Acknowledge All Alarms] applies to all unacknowledged alarms and affects all users. It shows the count and asks for confirmation once more.
Hide · Suppress · Out of Service
These three are determined the moment they are triggered. If applicable, the occurrence is recorded separately as a hidden occurrence rather than an alarm row, and it does not appear in the alarm list, notifications, alarm counts, or alarm performance aggregates. Alarms that were already active remain so even if hiding or suppression begins.
| Handling | Display Text | Who | Duration | Safety/Environment Alarms |
|---|---|---|---|---|
| Operator Hide | Temporary Hide | ADMIN in Advanced Alarm | Fixed time (1 min to 7 days). Automatically released when finished | Cannot be hidden |
| Planned Maintenance Suppression | Design Suppression | Automatic if a planned maintenance calendar is linked to the equipment | Until that calendar interval ends | Not suppressed |
| Repair Suppression | Design Suppression | Automatic if there is an ongoing breakdown repair on the equipment | Until repair is completed | Not suppressed |
| Out of Service | Out of Service | ADMIN in Advanced Alarm | No automatic release. Returns after return inspection | Can be removed, but a memo is mandatory |
- If two or more overlap at one moment, only one is recorded in the priority order: Out of Service > Temporary Hide > Design Suppression.
- Design suppression only applies to the equipment linked to that tag. It does not propagate to sub-equipment.
- If the condition is still active when hiding or suppression ends, the hidden occurrence closes (
SUPPRESSION_ENDED) and a new alarm occurs at that time. A notification is also sent then. Occurrences that turned off before the period ends do not appear in the alarm list. - Out of Service is different. An alarm that was active when removed is closed (release reason «Out of Service»). This does not mean the problem was resolved.
중복 알람 억제 differs from these three. It only reduces notifications (push) of the same type for a set time, while the alarms are still recorded.
Counts hidden by repair are viewed in the equipment maintenance tab as «N alarms hidden by this repair», and occurrences during temporary hide or out of service are viewed via «View occurrences during hide» in the Advanced Alarm card. For API, it is GET /api/v5/alarm/suppressed, and for MCP, it is list_alarm_suppressions.
Screen Structure — 5 Sub-menus
| Screen | Primary User | Summary |
|---|---|---|
| Alarm | Operator | Real-time alarms + Statistics + Ranking (Main Monitoring) |
| Alarm Search | Operator/Analyst | Precise search by range + condition |
| Advanced Alarm | ADMIN | Rule definition, deployment, release, redeployment, temporary hide, out of service |
| Alarm Analysis | Analyst | 7 types of visualization charts (ranking, time slot, ratio, trend, treemap, etc.) |
| Alarm Performance | Manager/Analyst | Alarm system performance evaluation (ISA-18.2 / EEMUA 191) |
All five menus are visible to ADMIN, API, and USER. USER can view the list in Advanced Alarm, but write buttons (Add, Deploy, Release, Redeploy, Edit, Delete, Hide, Out of Service) are not displayed.
Alarm Screen (/alarm/index)
Top Search Area
🔔 알람 [기간] [사이트 ▼] [☐ 신호 끊김만] [검색] [알람 검색] [↻]
| Control | Description |
|---|---|
Period (alarm_time) | Start ~ End |
Site (site_id) | Site name [Description] |
| Signal Loss Only | Keeps only alarms where the state is unknown due to signal loss in the list. Statistics remain based on the overall search |
| Search | Reads again with current conditions |
| Alarm Search | Navigates to the Alarm Search screen |
| Refresh | Reads again with current conditions |
Timeline
| Item | Description |
|---|---|
| Header | 🕐 "Timeline" + Display range label (alarm_timeline_range_label) |
| Quick Time Range | 6 buttons (10M/30M/1H/6H/12H/ALL) — Same as other screens |
| Acknowledge All Alarms | ✓ — Visible only to ADMIN. Acknowledges all unacknowledged alarms (Acknowledgment) |
Timeline Area (alarm_timeline) | Alarm markers displayed by color on the time axis |
| No Data | "No data" guidance (alarm_timeline_no_data) |
Marker colors are based on priority (Status Code): Information (Green)/Warning (Orange)/Critical (Red).
Left — Recent List Table
Left col-lg-9 area. Alarms for the search period are shown in pages by latest first, with the total count written in the info line below the table. Pagination is available up to the latest 10,000 records.
| Column | Width | Description |
|---|---|---|
| 🔔 | 30px | Alarm bell icon |
| Alarm Number | 100px | Domain alarm ID |
| Importance | 80px | INFO/WARN/ERROR color badge |
| Alarm Start/End Time | 160px | Two lines — Occurrence/Release time |
| Tag | Location | 380px | Tag ID + Location (Asset path) |
| Message | Auto | Alarm message body |
| Acknowledge | 60px | Indicates if acknowledged |
Clicking a row navigates to the Alarm Detail screen. Acknowledgment is performed on the detail screen.
Right — Statistics/Ranking Panel
Two panels in the right col-lg-3 area.
Statistics Panel
| Item | ID | Description |
|---|---|---|
| Headline | alarm_count_label | "Searched Alarms : N cases" large text |
| Trend Sparkline | alarm_trend_spark_line (150px) | Alarm occurrence trend for the search period |
| Priority Donut | alarm_priority_count_chart (150px) | Information/Warning/Critical distribution |
| ON/OFF Donut | alarm_onoff_count_chart (150px) | Active (is_on=Y) / Released (is_on=N) distribution |
| Alarm Analysis (Button) | — | Top right → Navigates to Alarm Analysis screen |
| Empty State Guidance | stats_no_data | "No data" |
Ranking Panel (3 Tabs)
| Tab | ID | Display |
|---|---|---|
| By Tag/Alarm Type | rank_container | Ranking of tags/types with high alarm frequency |
| By Connection | rank_opc_container | Alarm frequency ranking by OPC/Edge unit |
| By Asset | rank_asset_container | Alarm frequency ranking by asset (equipment) unit |
If there is no data in a tab, a "No data" notice is displayed.
Alarm Help
Guidance box at the bottom of the screen.
❓ Alarm Help Monitor alarm messages received in real-time.
- The list shows alarms for the search period in pages by latest first, and the total count is written in the info line below the table.
- To perform a detailed search of all alarms by period/tag/priority, use the [Alarm Search] button.
- To add complex condition alarms based on EQL (Event Query Language), use [Add Alarm Setting] in the [Advanced Alarm] menu.
Alarm Search Screen (/alarm/advsearch/index)
🔔 알람 검색 [알람 기간] [← 알람 이동] [↻] [☰ ▼]
There is a Range + Condition search card on the left, and a Statistics Chart + Result List on the right.
Page Header
| Button | Action |
|---|---|
Alarm Period (alarm_time) | Date-time range |
| ← Move to Alarm | Navigates to the Alarm screen |
| ↻ Refresh | Refreshes the screen |
| ☰ ▼ | Dropdown — Navigates to other alarm screens |
Left — Search Card
The range is narrowed in the order of Site > Area > Line > Equipment. Conditions are as follows:
| Field | Option |
|---|---|
Importance (priority) | All / Information (INFO) / Warning (WARN) / Critical (ERROR) |
Active (is_on) | All / Active / Ended |
Acknowledge (acknowledged) | All / Acknowledged / Unacknowledged |
Content (description) | Keywords for the alarm message body |
Clicking Search below the card displays results on the right.
Right — Results
The ✓ Acknowledge All Alarms at the top of the result list is visible only to ADMIN and acknowledges all unacknowledged alarms, not just the search results (Acknowledgment).
Statistics Chart Area (Above list, displayed only when results exist)
| Area | Content |
|---|---|
| Priority Donut | INFO/WARN/ERROR share donut |
| Time Series Spark | Sparkline of alarm occurrences during the search period |
Result Table (6 Columns)
| Column | Width | Display |
|---|---|---|
| 🔔 | 30px | Bell icon |
| Alarm Number | 100px | Alarm serial number — Click to move to Alarm Detail screen |
| Importance | 80px | Information/Warning/Critical color badge |
| Alarm Date | 160px | Occurrence time |
| Tag Name | 200px | Occurring tag |
| Content | Auto | Alarm message body |
Alarm Analysis Screen (/alarm/analysis/index)
This screen analyzes alarm data using various visualization charts.
| Number | Position | Action |
|---|---|---|
| ① | Left Range | Select Site · Area · Line · Equipment («All …» means everything below it) |
| ② | Left Condition | Visualization selector — Choose one chart to draw in the results |
| ③ | Bottom Left | Analyze — Results remain empty until pressed |
| ④ | Screen Header | Alarm Period — Default is one month |
| ⑤ | Body | Results — The chosen chart is drawn |
The figure above shows this state — 「Please select the alarm period and visualization, then execute analysis」.
7 Types of Visualization — Options in the selector
There are seven options in the ⑤ Visualization selector. The names below appear exactly as seen on the screen.
| Screen Name | What it shows |
|---|---|
| Frequency Rank by Tag | Ranking of tags with the most alarms |
| Distribution by Time Slot | Occurrence frequency by 24-hour time slot |
| Ratio by Priority | Ratio of Information / Warning / Critical |
| Daily Trend | Occurrence trend by date |
| Daily Ratio Calendar | Priority ratio for each day in the calendar cells |
| Time Pattern by Day | Day × Time slot intersection — used to find repeating time slots |
| Equipment Treemap | Alarm distribution by area in the asset hierarchy |
The screens matrix · eventdrops · heatmap still have server paths and drawing code, and the branch code (analysis-index.js) also handles them, but they are not in the visualization selector options. In other words, they cannot be selected from the current screen — these are the 「Day/Time Heatmap」, 「Day/Time Matrix」, and 「EventDrops」 mentioned in older documents.
(A limit of up to 30 tags remains in the code for the heatmap path — 「Up to 30 tags can be selected for the calendar heatmap」.)
Combined Filters
| Control | Options |
|---|---|
| Priority | All · Information · Warning · Critical |
| Branch | Hourly · Daily · Monthly · Yearly — Time aggregation unit for the chart |
Usage Procedure
- Check tags to analyze in the left tree — if none are selected, it is blocked with 「Please select tags to analyze」
- Set Alarm Period, Visualization, Priority, and Branch
- Press Analyze (results are empty until then)
- Narrow down the specific time or asset where a pattern is seen using Alarm or Alarm Search for detailed inquiry
- If threshold adjustment is needed, modify band alarms in the alarm settings of Tag, and EQL alarms in the Advanced Alarm screen
Alarm Analysis is suitable for post-operational pattern analysis. For real-time monitoring, use the Alarm Screen.
Alarm Performance Screen (/alarm/performance/index)
This screen evaluates whether the alarm system is at a level that operators can handle using ISA-18.2 / EEMUA 191 metrics. Select Site · Area · Line from the left card and press Inquire (no equipment-level unit). The period is set in the header as Last 7 Days · Last 15 Days · Last 31 Days.
| Number | Position | Action |
|---|---|---|
| ① | Left | Inquiry Conditions — Site · Area · Line, Inquire |
| ② | Screen Header | Period — Last 7, 15, or 31 Days |
| ③ | Top of Body | 8 Metric Tiles |
| ④ | Below that | Alarm count trend (10-min bars, EEMUA standard dashed lines) |
| ⑤ | Bottom | Priority distribution · Top 10 frequent alarms · Chattering alarms · Stale alarms |
Metrics only count occurrences actually notified to the operator. Occurrences opened during temporary hide, design suppression, or out of service are excluded and appear separately in the Hidden/Suppressed tile.
8 Metric Tiles
| Tile | Meaning |
|---|---|
| Avg Alarm / 10 min | Total notified occurrences in period ÷ Number of 10-min windows in period (including windows with no alarms). Evaluated by the band below |
| Max 10-min Alarm Count | Occurrence count and start time of the busiest 10-min window |
| Alarm Flood Time Ratio | Ratio of windows that exceeded 10 cases. Exactly 10 cases is not considered a flood |
| Top 10 Alarm Share | The percentage of the total alarms accounted for by the 10 most frequent alarm settings |
| Chattering Alarm Settings | Number of settings that turned on for the third time or more within 1 minute and the number of such occurrences |
| Stale Alarms | Number of alarms that have been active for more than 24 hours |
| Hidden/Suppressed | Current number of temporary hide, design suppression, and out of service |
| Notified Alarm Count | Total occurrences in period (excluding those during hide/suppression) |
| Avg Alarm / 10 min | Evaluation |
|---|---|
| ≤ 1 | Very Good |
| ≤ 2 | Manageable |
| ≤ 10 | Overloaded |
| > 10 | Flood |
Charts and Tables
| Card | Content |
|---|---|
| Alarm Count Trend | 10-min bars (if period exceeds 7 days, 1-hour bars — height is the 10-min average for that hour). Dashed lines 1, 2, and 10 follow the criteria above, and red bars indicate floods. Clicking a bar navigates to Alarm Search for that time |
| Priority Distribution | Actual ratio compared to ISA-18.2 recommendation (High 5 · Medium 15 · Low 80 %). Critical, Warning, and Information correspond to High, Medium, and Low |
| Top 10 Frequent Alarms | Alarm setting · Tag · Priority · Occurrences · Ratio |
| Chattering Alarms | Up to 50 rows, sorted by frequency |
| Stale Alarms | Up to 100 rows, sorted by oldest |
If there are no alarms in the period, numbers are 0, the ratio is -, and «No alarms in this period» appears on the chart.
The same values are also read via GET /api/v5/alarm/performance?scope_type=SITE|AREA|LINE&scope_id=…&days=1..31 (or from · to, up to 31 days) and MCP get_alarm_performance. The screen, API, and AI use the same calculation.
Advanced Alarm Screen (/alarm/config/index)
🔔 고급 알람 [← 알람 이동] [↻] [▼] [➕ 알람 설정 추가]
This screen manages alarm settings. EQL rules are defined to issue alarms under arbitrary conditions, and settings can be temporarily hidden or taken out of service. (Similar to CEP Triggers, but dedicated to alarm issuance.)
Band Alarms and EQL Alarms
| Band Alarm | EQL Alarm | |
|---|---|---|
| Count per Tag | One | Multiple |
| Creation Point | Tag Alarm Setting — Created by Platform when tag is saved | Add Alarm Setting on this screen |
| Name | Automatic — <태그 이름> 밴드 알람 (in installation language) | Defined manually |
| Edit Point | Tag Alarm Setting. Editing a row navigates to the Tag screen | Alarm Setting Form |
If a band alarm is created or modified like an EQL alarm, it is rejected with «Band alarms are changed in the tag's alarm settings» (HTTP 409, alarm.band.edit_on_tag). A DB uniqueness constraint ensures only one band setting per tag.
Alarm Setting ID
New alarm setting IDs are issued by the server — ALARM_CONFIG_ + 6 digits (e.g., ALARM_CONFIG_000376). Band alarms follow the same format.
- Older formats (
ALARM_CONFIG_00375,ALARM_CONFIG_TAG_…) remain valid but can only be edited. - When creating via API, use the ID received through
GET /api/v5/alarm/config/new-id. Manually defined IDs are rejected with HTTP 422alarm.invalid.alarm_config_id. - Do not use Tag ID to create or search for Alarm Setting IDs. Band settings for a tag are found where
GET /api/v5/alarm/config/list/by-tag?tag_id=…isalarm_config_priorityandBAND.
Like band alarms, EQL alarms are linked to one tag — the Target Tag in the form is mandatory. To set an alarm on a condition combining multiple tags, first create that combination as a Virtual Tag and set the alarm on the virtual tag.
If an EQL alarm reads other tags, mixes multiple tags, or lacks a tag condition, saving is rejected, and [Execute EQL] in the form is also rejected for the same reason(alarm.eql.one_tag, HTTP 422) — see the Virtual Tag section for detailed phrasing and handling of already saved alarms.
List Table
The ADMIN list shows all alarm settings. Therefore, alarms created by others and band alarms generated by tag saves can be hidden or taken out of service here. Lists for other roles show only the alarm settings they registered. All alarm settings for a single piece of equipment are viewed in the alarm tab of the asset tree.
Three buttons are located on the right of the list panel header: Deploy Selected Alarm Settings / Release / Redeploy All.
| Column | Width | Description |
|---|---|---|
| Select | 50px | Checkbox for batch deployment/release |
| Status | 80px | Deploy/Release badge |
| Alarm Setting ID | 150px | Server-issued ID — Alarm Setting ID |
| Alarm Setting Name | Description | 200px | Operator meta |
| Importance | 80px | Priority of the alarm to be issued (INFO/WARN/ERROR). For band alarms, this is BAND — the actual priority is determined by the band that triggered |
| Message | Auto | Message body upon alarm issuance |
| Total Executions | 70px | Cumulative matching count |
| Last Occurrence | 110px | Most recent occurrence time |
| Error | 70px | Cumulative error count |
| Reg. Date | 110px | |
| Last Modified Date | 110px | |
| Action | 110px | Eye icon (Details) + ⋯ Menu: Edit · Temporary Hide · Out of Service · Delete |
Clicking the name or ID of a row navigates to the alarm setting details (/alarm/config/view/{id}), where you can view occurrences over the last 7 days, change history, and resolution guides. In the change history, band settings for a tag are grouped by tag.
Temporary Hide
ADMIN can hide new occurrences of a setting for a set period using the Temporary Hide option in the ⋯ menu of a list row.
| Input | Value |
|---|---|
| Hide Time | 30m · 2h · 8h · Manual entry (1–10080 min, max 7 days) |
| Reason | Maintenance · Sensor Failure · Other |
| Memo | Optional. Max 500 characters |
Hidden settings appear in the Temporary Hidden Alarms card below the list — Reason · Memo · Hidden by · Hidden time · Time remaining · Occurrences during hide. They automatically reappear when the time expires; to release early, click Show Again on that row. Records during the hidden period are preserved.
Safety/Environment alarms are rejected with «Safety/Environment alarms cannot be hidden».
Out of Service and Return
For periods where the alarm itself becomes meaningless, such as sensor removal, equipment demolition, or calibration, ADMIN removes it via the Out of Service option in the ⋯ menu of a list row.
| Input | Value |
|---|---|
| Reason | Sensor Removal · Equipment Demolition · Calibration · Modification · Other |
| Memo | Mandatory if reason is «Other» or for Safety/Environment alarms. Max 500 characters |
| Scheduled Return Date | Leave blank if not set. It does not return automatically after the date passes |
Removed settings appear in the Out of Service Alarms card — Reason · Memo · Approver · Start · Scheduled Return · Elapsed days · Occurrences during exclusion.
Return is performed via the Return button on that row. It must pass a return inspection — the current value must be within the normal range and a value must have entered within 10 minutes after the exclusion.
| Result | Outcome |
|---|---|
| Pass | Returns to service |
| Not in normal range | «Current value … — Not returned as it is not yet in the normal range.» Out of service is maintained, and diagnosis is triggered |
| No value · Stale value | «Not returned because no value entered after exclusion or the last value is older than 10 minutes.» Out of service is maintained, and diagnosis is triggered |
| Return Anyway | Returns by providing a reason, bypassing the inspection. Any active conditions immediately become alarms |
| Diagnosis | When |
|---|---|
diagnostic.alarm.out_of_service_long | Out of service exceeds 30 days or the scheduled return date has passed |
diagnostic.alarm.return_check_failed | Return inspection failed |
Alarm Setting Form
Navigate to the form screen (/alarm/config/form) via the top Add Alarm Setting button or a row action.
| Number | Position | Action |
|---|---|---|
| ① | Top Right | [Execute EQL] — Validates the sentence in ③ using the same rules as saving without actually saving — See below |
| ② | Basic Info | Alarm Name · Description · Importance (Info · Warning · Critical) — All mandatory |
| ③ | Condition (EQL) → EQL | Alarm condition sentence. As per the help «Alarm occurs when a result matching the EQL condition exists», it must create a band column and read only one target tag. If rejected, the cause appears in the box below this editor |
| ④ | Message | Alarm message. Substitution tokens {tag_id} · {tag_name} · {timestamp} · {value} |
| ⑤ | Target Tag | Mandatory — «Please select a tag name to be associated with the set alarm.» |
| ⑥ | Duplicate Alarm Suppression → Suppression Time (min) | «Suppresses alarm pushes of the same type for the set time.» |
| ⑦ | Detection Setting (ISA-18.2) → On Delay | «Alarm sounds only if the condition persists for this time (ms). Set to 0 to disable, max 3,600,000 ms (1 hour).» |
| ⑧ | Bottom Right | [Save] |
| Input | path | Description |
|---|---|---|
| Alarm Setting ID | alarm_config_id | Read-only (Server issued) |
| Alarm Setting Name | alarm_config_name | Human-identifiable name |
| Description | description | Memo for rule purpose |
| EQL | epl | Alarm matching condition. Must create band column and 'NORMAL' and read only one target tag (e.g., SELECT timestamp, tag_id, tag_name, value, quality, error_code, CASE WHEN to_double(value) > 80 THEN 'HIGH' ELSE 'NORMAL' END AS band FROM Point(tag_id = 'TAG_…').win:length(1)) |
| Target Tag | tag_id | Mandatory. One tag this alarm is linked to — Screen help «Please select a tag name to be associated with the set alarm.» |
| Importance | alarm_config_priority | INFO / WARN / ERROR |
| Message | message | Substitution tokens {tag_id} · {tag_name} · {timestamp} · {value} |
| Suppression Time (min) | duplicate_check_time | Suppresses alarm pushes of the same type for this duration |
Upon saving, it is immediately deployed to the engine. If the engine rejects the sentence, the save is rolled back («The engine rejected the sentence, and changes were rolled back»). To stop it, check the item in the list and select Release; to restart, press Deploy Selected Alarm Settings.
[Execute EQL] — Validated by the same rules as saving
① [Execute EQL] views the sentence using the same rules as when saving. Results appear in the guidance box immediately below the EQL editor. The same location is used when saving (⑧) is rejected. This rejection does not appear as a separate toast.
| Case | Box | Content |
|---|---|---|
| Reads two or more tags · No tag condition · Joins with other tags | Orange «This EQL does not meet the rules» | «EQL alarms read only one tag — the tag of the alarm. …» |
No band column | Orange, same title | «EQL sentence must have a band column (... END AS band).» |
band cannot produce 'NORMAL' | Orange, same title | «EQL sentence must be able to produce band 'NORMAL'. Otherwise, the alarm will not turn off.» |
| Cannot be compiled | Red «Cannot read EQL» | Location chip «Line N Column M» (click to move cursor), cause sentence, fix hint, collapsed View Engine Original. A mark also appears at that position within the editor |
| Pass | Green «Validation Passed» |
All rejections are HTTP 422, and the mark is cleared once the sentence is corrected. If CEP is not running, it is not a sentence error; a toast appears instead of a box: «Unable to connect to CEP engine, so EQL could not be validated».
If a saved alarm cannot be loaded into the engine during server boot or redeployment, a diagnosis «Tag Alarm Deployment Failed» (diagnostic.tag.alarm_deploy_failed) is triggered for that alarm. The «Open Fix Screen» in the diagnosis details opens this form, and the diagnosis is automatically cleared once it is corrected and saved as deployed — Virtual Tag Operation.
EQL alarms can also be automatically generated using the
flow_create_alarm_confignode in Flow.
Alarm Detail Screen (/alarm/view/{alarm_seq})
Clicking an alarm in the Alarm or Alarm Search screens enters the detail screen. This is an analysis view that gathers one alarm and its surrounding context (tag, data, other alarms, statistics, recipients) on a single screen.
Top — Identification Header
🔔 알람 | {alarm_seq} [ 사이트 > 구역 > 라인 > 설비 > 태그명 ]
| Area | Action |
|---|---|
| ← Previous Screen | back() — Back to the previous alarm list screen |
| Refresh | refresh() — Reload screen |
Panel 1 — Alarm Information (panel 알람 정보)
A single-row table containing core information for the alarm. An Acknowledge Alarm button is in the panel header (all roles), and ADMIN also sees Acknowledge All Alarms.
Acknowledge Alarm acknowledges only this single alarm. If the alarm is already acknowledged, the button is disabled, and hovering over it shows who acknowledged it and when. Simply opening the screen does not constitute acknowledgment — Acknowledgment.
| Column | Width | Description |
|---|---|---|
| Importance | 200px | INFO/WARN/ERROR color badge |
| Active | 60px | Whether active |
| Alarm Start Time | 120px | Occurrence time |
| Alarm End Time | 120px | Release time (displays - if active) |
| Tag | 400px | 🏷 Tag name — Click to move to Tag Details |
| Message | Auto | Alarm body |
| Acknowledge | 100px | Indicates if acknowledged |
Data Situation fieldset
Displays the tag data in table format for the alarm occurrence time ± the inquiry period. A Move to Trend (goStorage()) button is on the right of the header.
| Column | Width | Description |
|---|---|---|
| Timestamp | 180px | Measurement time |
| Value | 200px | Tag value at that time |
| Unit | 50px | Measurement unit |
| Data Type | 100px | INT/FLOAT/BOOL, etc. |
| Quality | 100px | Data quality code |
| Error Code | 100px | Error at that time (if any) |
| Delay [MS] | 100px | Ingress delay |
| Attribute | Auto | Additional attributes |
You can see tag values immediately before and after the alarm occurrence and use Move to Trend to analyze longer time series.
Alarm List fieldset
History of other alarms for the same tag over the same period (${date_from} ~ ${date_to}). The Move to Alarm (goAlarm()) button on the right returns to the alarm menu.
| Column | Width | Description |
|---|---|---|
| 🔔 | 30px | Alarm bell icon |
| Alarm Number | 100px | Alarm ID |
| Importance | 80px | INFO/WARN/ERROR |
| Alarm Date | 160px | Occurrence time |
| Content | Auto | Alarm message |
| Acknowledge | 60px | Indicates if acknowledged |
| Acknowledge Time | 160px | Time of acknowledgment |
This table allows you to identify at a glance if other types of alarms occurred for the same tag during the same period.
Alarm Trend Panel
Visualizes the alarm occurrence trend during the period using time bars. It shows occurrence frequency and patterns (e.g., concentration at specific time slots).
Alarm Statistics Panel
Summary of alarm occurrence statistics. Counts by priority and ON/OFF status are displayed as donuts or bars.
Alarm Recipient Panel
Shows acknowledgment status and destination together.
- Top row — If Acknowledged, shows the acknowledger (user or employee) and time; if Unacknowledged, shows the [Acknowledge] button (only if acknowledgment permission exists).
- Table below — Actual sending records.
| Column | Description |
|---|---|
| Recipient | User or employee |
| Channel | Console · Mail · SMS · App |
| Sent Time | Time the notification was sent |
If not sent to anyone, «Not sent to anyone — was within a duplicate window or no recipients existed» is displayed.
Tag Information Panel
Detailed information for the tag where the alarm occurred — Settings, Alarm Band, Validation, OPC Mapping, Meta. For detailed items, refer to Tag Details.
Usage Patterns
| Analysis Flow | Screen Flow |
|---|---|
| Causal Analysis of Alarms | Alarm Info → Data Situation Table (Trend just before) → Move to Trend (30-min time series) |
| Tracking Repeating Alarms | Alarm List fieldset → Other alarms for same tag → Identify pattern |
| Threshold Review | Tag Info Panel → Check Alarm Band settings → Recalculate threshold via Predictive Analysis |
| Recipient Verification | Alarm Recipient Panel → Who received and who acknowledged |
Utilization Scenarios
| Scenario | Screen | Procedure |
|---|---|---|
| Check current active alarms | Alarm | Enter page → Top right bell badge + ON status alarms in left list |
| Analyze ERROR from yesterday | Alarm Search | Period = Yesterday / Importance = Critical / Select Site |
| Alarm pattern by day/time | Alarm Analysis | Identify repeating days/times in Time Pattern by Day |
| Adjust threshold for repeating alarms | Tag / Advanced Alarm | For band alarms, Tag Alarm Setting; for EQL alarms, modify rule → Redeploy |
| Evaluate alarm load | Alarm Performance | Avg Alarm / 10 min band → Top 10 → Organize chattering settings |
| Remove alarms during sensor calibration | Advanced Alarm | ⋯ → Out of Service (Calibration) → Return upon completion |
| Identify alarm bursts by equipment | Alarm Analysis → Equipment Treemap | Large area asset → Alarm Search for that asset only |
| Automate alarm notifications | Flow | flow_on_alarm Trigger → Send Email/SMS |
FAQ
Q. How many records are shown in the Alarm screen list? A. Alarms for the search period are shown in pages, and pagination is available up to the latest 10,000 records. For older records, narrow the period or use Alarm Search.
Q. Do alarms disappear when I press "Acknowledge All Alarms"? A. They do not disappear. Only the status changes to acknowledged, and active alarms remain active. However, it applies to all users — acknowledgment is per alarm, not per person.
Q. I am a general user (USER), and the [Acknowledge All Alarms] button is missing. A. Acknowledge All is for ADMIN only. USER acknowledges alarms one by one using [Acknowledge Alarm] in the alarm details. Field employees can also acknowledge via the field app.
Q. Alarms for equipment under repair are not sounding. A. If there is an ongoing breakdown repair on the equipment, newly triggering alarms for that equipment's tags are suppressed (except Safety/Environment alarms). Check «N alarms hidden by this repair» in the equipment maintenance tab. If still active after repair is completed, a new alarm occurs at that time.
Q. An alarm closed, but the release reason is «Out of Service». A. It was closed because a manager took that alarm setting out of service. This does not mean the cause was resolved.
Q. I want to automatically send Email/SMS when an alarm occurs.
A. In the Flow screen, you can automate this by connecting a flow_on_alarm trigger with flow_send_email / flow_send_sms nodes.
Q. I registered an EQL alarm, but it is not being issued. A. Check if the Status in the list is «Deployed». If it is «Released», check the alarm setting and press Deploy Selected Alarm Settings. If deployed, check if it was hidden via the ⋯ menu (Temporary Hidden Alarms / Out of Service Alarms cards) or if the equipment is under planned maintenance or repair.
Q. What does the area represent in the Treemap of Alarm Analysis? A. It is the cumulative frequency of alarms that occurred for that asset/tag. A larger area indicates an asset with more alarms.
Q. Alarm priority colors seem slightly different across screens. A. Only the subtle tones vary by area; the meaning is identical. Refer to Status Code Definition for accurate color mapping.
Q. Which variables can I use in the alarm message template?
A. Use the substitution tokens {tag_id} · {tag_name} · {timestamp} · {value} found in the form help.
Q. How can I keep alarm issuance history for a long time?
A. You can increase the default retention period or use a flow_on_alarm trigger in Flow to automatically load data into an external system.
Q. Why is the ON/OFF donut always 100% ON? A. This happens when there are no released alarms in the search results. This means all alarms in the result occurred and remained active without being released.
Q. What can I see in the Alarm Detail screen? A. You can see tag data at the time of occurrence, other alarms for the same tag, trends, statistics, acknowledgment status and destination, and tag information all together.
Related Screens
- Dashboard — Today's alarms in KPI cards, alarm counts in diagnosis panels
- Daily Statistics — Comparative analysis of alarms by site, asset, and tag
- Data Point — Alarm display options in trend charts
- CEP — Higher-level concept of alarms (Event Rules)
- Virtual Tag — When setting alarms on conditions combining multiple tags
- Flow — Automatic alarm routing and notification automation
- Status Code Definition — Alarm priority colors
- Equipment — Today's alarm count (Info/Warning/Critical) on cards