Riyadh, Saudi ArabiaPredictive Maintenance
Condition and usage monitoring that moves maintenance off the calendar and onto the signal — so the failure becomes a scheduled job instead of a stopped line.
/ The shift /
Flip between the operation as it runs now and the operation once the data is joined up.
Which wastes money on healthy equipment and misses the machine that was never going to make it to the next service.
Vibration climbed for three weeks. Temperature drifted. Nobody was watching that specific signal on that specific machine.
Because nobody can attribute it to a cause with confidence.
/ How it works /
Each step runs on the shared IoT Cube foundation, so nothing here is a separate system to maintain.
Retrofit vibration, temperature and current sensors go on the assets that matter, and existing PLC and SCADA tags are read alongside them.
Signals are linked to the asset, its duty and operating state, and a baseline is built for that specific machine.
Models and thresholds spot deviation from baseline, estimate severity and suggest a probable cause such as bearing wear or misalignment.
A work order opens in your CMMS with the evidence attached, and parts and downtime are planned before the failure chooses the time.
/ What you get /
Models compare live vibration spectra, temperature and current against each machine’s learned baseline. Drift is ranked by severity and matched to likely faults such as bearing wear, imbalance or misalignment.
Machines, gensets, pumps and mobile plant share one asset hierarchy. Sensor data and PLC tags are stored per asset, with operating state used to avoid false alarms.
High-frequency vibration and acoustic data are processed on an edge gateway into features such as RMS and spectral bands, so only what matters crosses the network.
When a fault crosses a severity level, a work order is created in your CMMS or EAM with the asset, trend, severity and recommended check.
Detected faults and their expected progression are shared with procurement, so parts are ordered for the machines that need them rather than by average consumption.
Downtime events, causes, MTBF and early-detected faults are recorded per asset, giving reliability reviews a record instead of an estimate.
/ Use cases /
Detect a rising vibration signature on a pump or motor and schedule the repair for the next planned stop.
Watch coolant, load and run hours on remote gensets, and dispatch a technician before the site loses backup.
Turn fault codes and engine data from trucks and mobile plant into planned workshop jobs instead of roadside breakdowns.
Share detected faults with procurement, so spares are ready for the machines that need them when work begins.
/ Benefits by team /
Maintenance
Technicians spend time on machines that are drifting, and work orders arrive with the fault already described.
Operations
Stoppages move from surprise to schedule, so production and delivery plans are not decided by a failure.
Finance
Downtime and cost avoidance are recorded per asset, giving capital and maintenance budgets a defensible basis.
IT
PLC and SCADA data is read through an edge gateway outside the control loop, with read-only connections by default.
/ In practice /
Three moments where the platform earns its place. Scenarios are illustrative.
The signal
The model flags a bearing-wear signature and ranks it as medium severity.
A work order opens in the CMMS with the trend and recommended inspection.
Reliability schedules the bearing change for the next planned stop.
The outcome
The repair happens on the plant’s schedule, not the pump’s.
/ Technology /
Device-neutral at the bottom, open at the top, and sovereign all the way through.
Keep the hardware you already have
Any carrier, any protocol
Sovereign, multi-tenant, AI-neutral
Data lands where decisions are made
/ What you measure /
The indicators this solution produces continuously, per asset, per site and per contract.
Hours of stoppage not scheduled in advance, per asset, line and site.
Faults found from condition signals before they caused a stoppage.
Share of maintenance work orders raised from schedule or condition rather than breakdown.
Average operating time between failures, per asset or asset class.
Time from a condition alert to a work order being raised and assigned.
Share of condition-based jobs where the required parts were ready when work began.
/ Rollout /
We review failure history and criticality with your reliability team and select the machines where an early warning changes the outcome.
Retrofit sensors and PLC connections go onto the pilot assets, and baselines start building while threshold alerts run from day one.
Detected faults are inspected and confirmed with technicians, and work orders flow into the CMMS without manual re-entry.
More asset classes, sites and models are added on the same hierarchy and workflow. Weeks, not quarters, per wave.
/ Security & sovereignty /
Data stored and processed inside Saudi Arabia, on infrastructure you can name in a tender.
Designed to align with NCA cybersecurity controls and the Personal Data Protection Law.
Tenant isolation, role-based access and a full audit trail across users, actions and devices.
Deploy into a sovereign cloud, your private cloud or on-premises without becoming a different product.
/ Built on IoT Cube /
Products already running this solution on the shared foundation.
Built on IoT Cube
Saudi IoT and telematics hardware — GPS tracking, RFID, AI cameras, sensors and RTLS, installed and supported.
Built on IoT Cube
On-demand car care — convenient vehicle maintenance and automotive services, booked in a few taps.
Built on IoT Cube
All-in-one logistics for transportation management, fleet maintenance, freight forwarding and customs operations.
/ Industries /
Industry
Machines, utilities and mobile plant reporting condition before they stop.
Industry
Remote assets and contractor movements on one sovereign layer.
Industry
Networks that report their own losses.
Industry
Continuous cold chain evidence, produced as a by-product of the operation.
/ Questions /
Our machines are old and have no sensors.
Most do not. Retrofit condition monitoring is normal and it is usually where we start.
Do we need a data scientist?No. The models ship with the solution, and you can bring your own if you have them.
How long before it is useful?It needs a baseline period to learn normal. That is weeks, not quarters, and threshold alerting works from day one.
Can it read from our PLCs and SCADA without disturbing control?Yes. Connections are read-only by default over OPC UA or Modbus, through an edge gateway that sits outside the control loop. Nothing is written back unless you decide it should be.
Which maintenance systems does it raise work orders in?Any CMMS or EAM with an API or import route. The work order carries the asset, the signal, the severity and the probable cause, so the technician starts informed.
Does analysis run on site or in the cloud?Either. High-frequency vibration can be processed at the edge and only features sent up, or everything can run in sovereign cloud. Industrial connections are designed to align with NCA controls.
/ Explore more /
Know where assets are, how they are performing, and where attention is needed.
Monitor conditions continuously and create the operational record while products move.
Connect journeys, behaviour, safety events and operational response.
Start with the problem. IoT Cube provides the foundation underneath.