Current plan
M-01 running
- Energy over 30 min
- 31.1 kWh
- Peak power
- 64.7 kW
Machine Monitoring
See every machine before someone reports it
The gateway reads status from the PLC or digital inputs, and values from sensors on the machine, and sends them to the cloud — so you see which machine is running, which has stopped, and which value is starting to drift.
Common problems
- You only learn when and how long a machine stopped once someone reports it
- Downtime and OEE figures come from spreadsheets filled in after the fact
- A machine starts running hot, vibrating or drawing more current and nobody sees the trend
What we measure
Values read from your equipment
Read through the gateway, shown on the dashboard, with history charts.
- Run / stop status
- Cycle count
- Temperature
- Motor current
- Vibration
Calculated indicators
Coming soonThese indicators are in development and will be enabled step by step. You can preview them in the demo.
- Runtime
- Downtime
- Availability
- OEE
Data path
From shop-floor equipment to dashboards and alerts.
Scroll to explore the connection path →
PLC / Sensor
The machine's PLC or added sensors
Modbus / DI
Modbus RTU / TCP or digital inputs
AIoT Gateway
Reads values; buffers when the connection drops
4G
Dual SIM or site LAN, encrypted with TLS
AIoT TH Cloud
Time-series storage, isolated per organization
Dashboard / Alerts
On the web; alerts via LINE and email
What you get today
Available as soon as the gateway is installed.
Dashboards
Latest values and charts for every data point, on desktop and mobile.
History
Detailed data for 30 days on every package; longer history is kept as summarised data, by package.
Threshold alerts
Set a high / low limit per data point. When a value goes out of range, you are alerted via LINE and email.
Offline alerts
Get alerted when a gateway or device stops sending data — no need to wait for someone to notice.
AI Insights
Coming soonAI capabilities are on our roadmap. They are not available yet.
- Anomaly detection based on each data point's usual pattern
- An AI-written daily summary report
- An AI assistant that answers questions from your data, citing the actual readings
Example: surface vibration or temperature trends that drift from that machine's usual behaviour.
FACTORY APPLICATIONS
From a site problem to a practical system
Illustrative system designs with equipment, data, responsible teams and project conditions. These are not customer case studies.
DIGITAL TWIN · SIMULATED
Simulated digital twin: site, equipment and events in one view
This simulated digital twin is a model running in your browser. It does not mirror a PLC or customer site and does not control equipment.
Loading 3D map…
Live (simulated) · 08:00:00
Illustrative 3D layout, not surveyed coordinates. Scene labels show values from the simulated twin. Select equipment on the scene or in the twin tree.
Total power62.0 kW
M-02 temperature46.0 °C
M-02 vibration2.10 mm/s
WHAT-IF · SIMULATED
Current plan vs alternative · next 30 simulated minutes
Alternative
M-01 stopped
- Energy over 30 min
- 17.4 kWh
- Peak power
- 36.2 kW
Computed from the browser simulation to compare choices; not a prediction for a real site.
Inspection and command history
Select a scene device, then inspect or control it
Data path
PLC / Meter → Modbus TCP / RS485 → Gateway → MQTT TLS :8883 → Monitor
CCTV uses IP over PoE / NVR independently of LoRaWAN. This scene is illustrative and has no live video.
Operator view
Connect power readings and machine states to site locations for maintenance inspection.
ILLUSTRATIVE SYSTEMSee machine states and investigate events with history
Example system: a packaging line with several stations
Connect PLC or sensor signals so production and maintenance use the same observations before developing performance analytics.
This illustrative line has feeding, filling and packing stations. Start with run/fault states and counters exposed by the PLC. Existing controls are not automatically modified, and an offline gateway is not treated as evidence of a machine stop.
Problems and practical approaches
Staff must walk the floor to see which machines are running
Shift leaders lack a shared view across multiple machines
Approach · Define readable PLC states and display them by machine with update timestamps
Stop records rely on memory or retrospective notes
Investigations begin without reliable times and source signals
Approach · Retain relevant states and readings to identify investigation periods; a person still confirms the actual cause
OEE is requested, but only a run signal is available
A calculated number may not represent production correctly
Approach · Separate acquisition from analytics and define shifts, counts, good units and ideal cycle times before agreeing OEE formulas
Equipment and components
- PLCs or sensors exposing agreed states
- A gateway with matching Ethernet, RS485 or digital-input interfaces
- Cloud assets separated by station
- Dashboard values and history shared by production and maintenance
Data to monitor
- Run, stop or fault values defined with the machine owner
- Production counters, temperature or current where a data source exists
- Gateway connectivity and update time, separate from machine state
Data journey
- PLC / sensor
- Readable values or states
- Gateway
- Cloud assets
- Production view
Workflow and responsibilities
Define machine states
Production and PLC owner
Agree run, stop and fault meanings, including changeovers and planned stops
Prepare a signal list
Automation engineer
Document addresses, types, units and a read method that does not interfere with control
Check actual events
Installer and operator
Compare displayed states with machine behavior during an agreed test
Monitor the shift
Shift leader
Check states and freshness, using history to select periods for investigation
Investigate on site
Maintenance
Inspect the machine and record causes through the factory process rather than assuming them from one signal
Review analytics inputs
Production engineering
Validate counts, shifts and formulas before adding downtime or OEE analysis
How each factory team benefits
Shift leaders and production
Coordinate using station-level readings and event timestamps
Maintenance
Use historical signals as evidence rather than relying only on recollection
Automation engineering
Maintain and extend the system using shared point definitions
Factory management
Identify available and missing inputs before investing in performance analytics
Who this is suitable for
- Factories with readable PLC values or accessible state signals
- Production and maintenance teams willing to agree state definitions
- Teams collecting reliable inputs before separately scoped downtime or OEE analytics
Acceptance checks
- Dashboard states match the agreed machine test cases
- Machine stops can be distinguished from gateway disconnections
- Machine owners accept the documented signal definitions
DETAILS & PLANNING
Agree machine-state definitions before measuring performance
A run signal does not explain every event
Interpret run, fault and cycle-complete signals with the machine owner. No-work conditions, planned stops and communication loss may produce different data patterns. Do not label every gap or inactive signal as downtime without an agreed definition.
- Document each address and state meaning
- Separate machine state from gateway connectivity
- Agree how to handle flickering signals and missing data
Inputs required for an OEE project
OEE needs planned production time, operating time, a reference production rate and good-unit counts. Status signals alone are insufficient. OEE and downtime analytics require confirmed feature availability and jointly agreed definitions.
- Define shifts and planned exclusions
- Validate total and rejected production counts
- Record the ideal cycle time for each product
Related product families
Extend machine monitoring with SCADA, factory control and OEE, using RFID or barcode to associate events with jobs, lots and products. Define production inputs and separate readings from control commands.
- Match device interfaces
- Define inputs and event workflows
- Confirm scope and validate on site
Systems related to this page
Review each system’s scope before opening it or using the example to plan your project.
Simulated data · No physical PLC commands
Factory Operations
Explore machine states, energy, simulated readings and alarms with a 3D scene and example event history.
Demonstration with simulated data
Demo Center
Explore energy, machine monitoring and OEE examples with a 3D factory to understand data and workflows before planning a project.
Questions about this topic
Must the PLC program be changed?
It depends on whether the required data is already exposed. The PLC owner should confirm addresses and access. Program changes must be a separately agreed scope.
Does monitoring replace interlocks or emergency stops?
No. Monitoring is not a safety control system, and Phase 1 does not write commands to field equipment.
See an example, or talk to our team
The demo uses simulated data from a sample plant. Our team can assess your actual equipment and site.