Vertical plant analytics

See your plant live. Change the maths yourself.

DynamicPlantMesh turns raw plant-floor signals into live dashboards, trends, KPIs and event intelligence — in the browser, in real time, on a historian that ships with the platform. It connects OPC-UA, OPC classic DA and MQTT v5 equipment, and puts self-service analysis and configuration in the hands of your own engineers.

Request a demo Explore the platform

DynamicPlantMeshCrushing & Milling overviewLIVE

Add panelLinked1h24h7d30d

TrendTRMill 3 — shift analysis08:00–12:00

Bearing vibration 5.74 mm/sMotor current 413.0 A

KPIKPIMill 3

Availability86.3%Target 90 %

Shift to date · −3.7 %

Circular gaugeCGCrusher 2 · load

0100

Motor load

Linear gaugeLGCrusher 2

Bin level68 %0100

Event timelineETMill 3 — stateLinked → TrendLinked

06:0010:0012:55Unplanned stop · 42 min

Event log gridLOGEvent log · Last 24 h

Time

Equipment

Event

12:13

Mill 3

Unplanned stop

11:20

Kiln 1

Planned stop

11:00

Conveyor CV-201

Planned stop

10:31

Pump P-14

Low flow

08:15

Conveyor CV-201

Belt slip

07:42

Crusher 2

Unplanned stop

Event property chartEPCEvent duration by cause · 7 d

Bearing vibration96
Tramp metal48
Refractory inspection45
Wet feed31
Belt inspection30
Suction blockage19

min

Illustrative view — replaced by a product screenshot when available

A process plant with OPC-UA, OPC classic DA or MQTT equipment
The platform ships with its own historian backend and connects the equipment you already have — nothing on the plant floor is replaced.
Engineers who change the KPI maths
Process and reliability engineers who own KPI, alarm and calculated-trace logic themselves — as expressions, not code releases.
On-premise, data-sovereign
The platform and its historian install on Windows hosts you already manage, with no container orchestration team required.

7dashboard panel types
19statistics per trace
40+expression functions
8MCP tools for external agents
OPC-UA · OPC DA · MQTT v5connectivity
6cooperating services

See everything live
Real-time streaming from field device to browser: dashboards, trends, gauges and event timelines that update as the plant runs.
Configure, don't code
KPIs, alarms and calculated series are authored as expressions in the app. Changing plant maths never needs a software release.
AI-ready by design
Built-in assistants, powered by DBS Invenio, answer plain-language questions about plant data, and an MCP server opens the historian to your own AI agents.

Trend statistics, not just lines
19 statistics per trace — from min and max with their timestamps to quartiles, skewness and RMS — generated from the trend on screen, so the spreadsheet round-trip disappears.
Event and downtime intelligence
How long, how often, what it looked like last time and what the process was doing while it happened. One timeline, one click from the event to the trend behind it.
Expression-authored KPIs
KPIs, alarm logic and calculated traces are expressions your engineers write in the app, with 40+ plant functions and no software release to change a calculation.
AI grounded in the live historian
In-app assistants answer from the live historian — the trend assistant from exactly what is on screen — and 8 MCP tools give your own agents governed access to equipment, events, KPIs and time-series.
Long, fine-grained history
“Show me this pump’s vibration during the same shift last quarter.” Time-series from per-second to per-day resolution over any historical range — the answer is a trend, not a ticket.

DynamicPlantMeshMill 3 · Live trendLIVE

LiveOffline1h24h7d30d4hStatisticsAsk AIExport PNGAnnotate

Vibration mm/sMotor current A · Belt speed m/s

2345673904004104204302.32.352.42.452.508:0009:0010:0011:0012:00Vibration rise 10:16Belt speed 2.42 m/sMotor current 402.7 ABearing vibration 3.16 mm/s

Bearing vibrationmm/s · axis 1
Motor currentA · axis 2
Belt speedm/s · axis 3

Illustrative view — replaced by a product screenshot when available

DynamicPlantMeshEquipment events · 06:00–14:00LIVE

Group byNoneEventEquipment

06:0008:0010:0012:0014:00

Crusher 2Conveyor CV-201Mill 3Pump P-14Kiln 1

12:55

Unplanned stop42 minlast occurrence 38 min (Tue)

Running
Unplanned stop
Planned stop
Warning
Idle
Completed
In progress
Now 12:55

Illustrative view — replaced by a product screenshot when available

Trends

From live signal to statistical insight

One trend workspace: watch samples stream in, overlay the events that matter, and turn the visible window into statistics without leaving the chart.

Live / Offline toggle with a pulsing LIVE indicator as samples stream in
Up to 13 Y-axes per trend, each with its own title
Overlay event windows side by side and drop crosshair annotations
19 statistics per trace, with the time of every minimum and maximum
Ask AI about exactly what is on screen
Excel export of resampled data and 2560×1440 PNG charts

13Y-axes per trend
19statistics per trace
1–48hquick-range menu

DynamicPlantMeshTrend analysis · Statistics

Window 08:00–12:00Export Excel

Summary statistics per trace
Trace	Count	Min	Max	Average	StdDev	RMS
▸ Bearing vibration (mm/s)	120	2.75	6.24	4.25	1.25	4.43
Motor current (A)	120	394.2	427.0	408.3	6.8	408.4
Belt speed (m/s)	120	2.34	2.46	2.40	0.03	2.40

Bearing vibrationMill3.Bearing.Vibration19 statistics08:00–12:00

Count: 120
Min08:58: 2.75mm/s
Max11:10: 6.24mm/s
Average: 4.25mm/s
Median: 3.43mm/s
StdDev: 1.25mm/s
Variance: 1.556
Range: 3.49mm/s
Sum: 509.82mm/s
Q1: 3.24mm/s
Q3: 5.68mm/s
IQR: 2.45mm/s
Skewness: 0.444
Kurtosis: -1.678
Coeff. of var.: 29.4 %
Mode: 3.32mm/s
RMS: 4.43mm/s
First value: 3.37mm/s
Last value: 5.74mm/s

Illustrative view — replaced by a product screenshot when available

Trends and live analysis on the platform page

Events

Every equipment event on one timeline

See what stopped, for how long and how often — then drill from the event to the trend that shows what the process was doing.

Group by equipment or event type; stack overlapping runs
In-progress and completed events colour-coded, with per-equipment overrides
Click an event to drill into its trend and event-log details
Compare selected events side by side
Chart event properties — stop duration by cause, for exampleNew
Export pivoted event grids to Excel or PDF

3timeline grouping modes
2export formats (Excel, PDF)

DynamicPlantMeshEquipment events · 06:00–14:00LIVE

Group byNoneEventEquipmentAuto-refresh−+Fit

06:0008:0010:0012:0014:00

Crusher 2Conveyor CV-201Mill 3Pump P-14Kiln 1

12:55

Unplanned stop42 minlast occurrence 38 min (Tue)

Running
Unplanned stop
Planned stop
Warning
Idle
Completed
In progress
Now 12:55

Illustrative view — replaced by a product screenshot when available

Event timelines on the platform page

KPI engine

Author KPIs and alarm logic as expressions

Plant calculations belong to the plant. Your engineers write and change them in the app — the same expression language spans KPIs, event logic and automated actions.

KPI and event trigger / reset logic as text expressions — no code deployment
40+ plant functions: historian aggregates, state durations, signal maths, bit decoding
Fire SMS, email, API calls or stored procedures from events
Evaluate per equipment, measure and shift, logging actual against target
Re-extract event logs over any historical range and watch progress live

40+expression functions
0redeploys to change a KPI

DynamicPlantMeshKPI configuration · Expression

Availability%Average motor currentA

1// Average motor current while running, ignoring idle samples under 50 A

2hiscalcavg('Mill3.Motor.Current', preveventstart('Shift start', 'Mill 3', currenttimestamp()), currenttimestamp(), 50)

3

4// Publish 0 until the mill has logged at least 10 running minutes this shift

5* if(stateduration('Running', 'Mill 3', preveventstart('Shift start', 'Mill 3', currenttimestamp()), currenttimestamp()) >= 10, 1, 0)

hiscalcavghiscalcavg('tag', start, stop, threshold)
hiscalcfirsthiscalcfirst('tag', start, stop)
hiscalclasthiscalclast('tag', start, stop)
hiscalcmaxhiscalcmax('tag', start, stop, threshold)

Function names are real; signatures shown are illustrative.

Illustrative view — replaced by a product screenshot when available

The KPI and event engine on the platform page

Inside the app, 2 assistants answer questions about your plant; an MCP server opens the same historian to agents you run yourself. Every answer is grounded in live plant data — the trend on screen, the configured events, the KPI log — not a generic summary.

Historian Agent — natural-language questions about plant data across historian connectors.
Ask AI — open it beside any trend and question exactly what is on screen; close the panel and reopen it while the trend is open and the thread is still there.
8 MCP tools — equipment, configured events with live trigger state, KPI definitions and logs, and time-series for agents that connect directly with an API key.
Powered by DBS Invenio — the assistants run through it, and it can be deployed with models hosted on your own infrastructure; ask about the local-model option for your deployment.
Fails closed — nothing activates until it is explicitly enabled and configured.

AI assistants and agent access

DynamicPlantMeshMill 3 · Ask AILIVE

3 items · 08:00–12:00Vibration mm/s

23456708:0009:0010:0011:0012:0010:16

Bearing vibrationmm/s
Motor currentA
Belt speedm/s

Ask AIPowered by DBS Invenio

You

Why did vibration rise after 10:15?

Assistant

Vibration before: 3.24 mm/s
Vibration after: 5.56 mm/s
Peak: 6.24 mm/s at 11:10
Motor current shift: +6.2 A

Before 10:15, Mill3.Bearing.Vibration averaged 3.24 mm/s and never exceeded 3.63 mm/s. From 10:16 it ramped to 5.56 mm/s on average — 72 % higher — and peaked at 6.24 mm/s at 11:10.

Mill3.Motor.Current moved from 405.6 A to 411.8 A over the same interval (+6.2 A), while CV201.Belt.Speed held at 2.40 m/s with a standard deviation of 0.03. The feed conveyor did not change, so the load step is on the mill itself.

Vibration and current rising together on a steady feed points to a mechanical load change at the drive-end bearing rather than a process upset. Check lubrication and bearing temperature next, and compare against the previous bearing-vibration stop on this mill.

Grounded in the visible windowCompare with the previous shiftShow motor current statisticsWhich events overlapped this window?

Ask about this trend…Send

Illustrative view — replaced by a product screenshot when available

Plant & production managers
Live dashboards, KPI tracking against targets and downtime visibility across equipment, shifts and sites.
Process & reliability engineers
Deep trend analysis, event comparison, a statistical toolkit and calculated traces for root-cause work.
Operations & control-room teams
Real-time equipment status, event timelines and alarm visibility with SMS and email notifications.
IT / OT integrators & data teams
Governed REST APIs, OPC-UA and MQTT v5 connectivity, native or Microsoft Entra ID sign-in, and AI-ready data access.

Field-to-dashboard data flow

OPC-UA devices → Field gateway service
OPC classic DA devices → Field gateway service
Field gateway service → MQTT v5 broker (live path)
Field gateway service → Store-and-forward buffer (while the broker is unreachable)
Store-and-forward buffer → MQTT v5 broker (replay on reconnect)
MQTT v5 broker → LIVE trends in the browser (live path)
MQTT v5 broker → Expression evaluation
Expression evaluation → KPI & event logs
KPI & event logs → Historian
Historian → Dashboards & timelines

Field signals arrive over OPC-UA and OPC classic DA and are published onto the embedded MQTT v5 broker, where the flow splits: a live path feeds LIVE trends directly in the browser, and a processing path evaluates expressions into KPI and event logs stored in the historian for dashboards and timelines. If connectivity drops, the store-and-forward buffer replays samples on reconnect, so outages become delays bounded by the configured disk buffer.

DynamicPlantMesh installs on premise, on Windows hosts, as 6 cooperating services around an embedded MQTT v5 broker and its own historian backend on a relational database. A field gateway service connects OPC-UA and OPC classic DA equipment, buffers to disk while the broker is unreachable and replays on reconnect — an outage becomes a delay, bounded by the configured disk buffer. Each service restarts and scales on its own, so a heavy KPI recalculation never touches the web app.

Connectivity and integration Architecture and deployment

Configure a plant solution — don't build one.

See DynamicPlantMesh running against plant data — and bring the KPI your engineers would change first.

Request a demo Explore the platform