Skip to content
Machvory Digital Engineering
IDEN

OPTIMUS · PREDICTIVE MAINTENANCE

Use condition data to decide when an asset needs attention

Condition trends, anomaly detection, and maintenance history combined into recommended actions, so the maintenance schedule follows asset condition and not just the calendar.

See the screens

Platform: OPTIMUS · Asset Reliability

OPTIMUS ML Anomaly Detection page showing data readiness per sensor category (temperature, kWh meter, vibration, CT current) with the number of devices and data rows collected so far
Captured from a running system, not an illustration.

The problem

Maintenance is scheduled by the calendar, not by condition

Interval-based preventive maintenance treats every asset the same. Some machines end up serviced too often, while the ones wearing out faster sit waiting for their next slot.

Signs we usually find

  • The same maintenance interval for assets under very different loads
  • A change in condition does not change the priority of the work
  • Small anomalies go unnoticed until they become a fault
  • No data to justify pushing a job back or bringing it forward

Used by

  • Reliability Engineer
  • Maintenance Manager
  • Plant Manager
  • Asset Management Team

Data it reads

  • Continuous condition data from condition monitoring
  • Alarm history and maintenance work history
  • How the asset runs: load, running hours, environment
A Machvory vibration sensor mounted on a machine bearing housing
A Machvory vibration sensor mounted on a machine bearing housing

How it works

From the machine to a decision

  1. 01

    Condition data

    Continuous telemetry

  2. 02

    Baseline

    Normal pattern per asset

  3. 03

    Anomaly detection

    Deviations flagged

  4. 04

    Recommendation

    Action & priority

The system

What it looks like on screen

OPTIMUS Analysis & Diagnosis page with trend analysis per utility, threshold limits, recent alarms and events, and live temperature and humidity readings
OPTIMUS Analysis & Diagnosis page with trend analysis per utility, threshold limits, recent alarms and events, and live temperature and humidity readings

Capabilities

What you can do with it

  • Condition trends

    You see which way condition is moving, not just the current value.

  • Anomaly detection

    Deviations from an asset’s normal pattern are flagged automatically.

  • History in context

    Alarms, past work, and running hours read side by side.

  • Action priority

    Assets ranked by risk, not by the order the reports came in.

  • Specific recommendations

    Findings turned into concrete proposals for an inspection or a repair.

  • Linked to work orders

    A finding can go straight into a scheduled job in the CMMS.

  • Post-repair check

    Condition after the repair compared against condition before it.

  • Adaptive thresholds per asset

    Normal limits follow each asset’s own behavior and operating point.

On the floor

How it is used day to day

  1. 01

    Anomaly

    Detected in the data

  2. 02

    Diagnose

    Engineer reviews it

  3. 03

    Work order

    Job is scheduled

  4. 04

    Execute

    Repair is carried out

  5. 05

    Verify

    Condition is rechecked

Impact

Why it matters

01

Maintenance work driven by condition

Assets in stable condition do not get torn down just because the calendar says so, and assets that are getting worse get attention sooner.

02

Repairs that are planned, not rushed

Planned work needs less downtime and does not force emergency spare parts purchases.

03

Decisions you can explain

Every proposed action comes with supporting data, so maintenance budget discussions run on evidence.

Start with this alone

Predictive maintenance needs condition data, so it usually starts alongside or after condition monitoring. The resulting work can run straight through the OPTIMUS CMMS.

In practice

Across different kinds of plants

Rotating Equipment
Flag a gradual rise in bearing vibration before it turns into a fault.
Utility
Review compressor condition before deciding on an overhaul.
Oil & Gas
Track condition changes on process equipment in critical areas.
Manufacturing
Sequence maintenance work across machines by their actual risk.

Integration

It connects to what you already have

No PLC, meter, or ERP needs replacing. Machvory reads from the equipment already installed in your plant.

  1. 01

    PLCs, sensors, meters

    Data at the machine

  2. 02

    Edge gateway

    Acquisition and buffering

  3. 03

    Machvory integration layer

    Modbus, OPC UA, MQTT

  4. 04

    TNS / OPTIMUS

    Analytics and workflow

  5. 05

    ERP, SAP, BI

    Enterprise reporting

  • Siemens, Mitsubishi, Omron, Allen-Bradley PLCs
  • Modbus TCP and Modbus RTU
  • OPC UA
  • MQTT
  • Power and energy meters
  • Vibration, temperature, current sensors
  • Existing SCADA and HMI
  • ERP and SAP
  • REST API and SQL

Next steps

What usually comes next

Start with one need and add the rest when you actually need it.

optimus · Asset ReliabilityCondition MonitoringVibration, temperature, current, and other condition parameters tracked continuously on critical assets, with thresholds, trends, and tiered alarms.Learn moreoptimus · Asset ReliabilityAsset Health MonitoringA health score per asset, a ranking based on condition and criticality, and a watchlist of critical assets that makes the order of work clear.Learn moreoptimus · MaintenanceCMMSPreventive and corrective maintenance, scheduling, work orders, checksheets, spare parts, and work history all tied to the same asset.Learn moreoptimus · MaintenanceMaintenance AnalyticsMaintenance KPIs, work order performance, preventive maintenance compliance, and repeat failures calculated from the work actually recorded in the system.Learn more

See the whole platform OPTIMUS

FAQ

Questions we get a lot

Does this replace preventive maintenance?

It does not replace it; it adjusts it. Mandatory and regulation-driven jobs stay on their schedule, while work that depends on condition can be brought forward or pushed back based on the data.

How much data do we need before it becomes useful?

Threshold-based alarms are useful from day one. Trend analysis and anomaly detection get sharper once the system has seen the asset run through its range of normal operating conditions.

Does this guarantee every failure gets predicted?

No, and we do not claim it does. What the system does is watch condition continuously, flag changes and anomalies early, and give enough context for the maintenance team to decide on an action before things turn critical. The final call stays with the engineer.

Do we have to replace our existing PLCs or machines?

No. Machvory reads from what is already installed over Modbus, OPC UA, MQTT, or a power meter. Where a machine does not expose the data point we need, we add a sensor or gateway to that machine alone rather than touching its controls.

Can it run on-premise, with no cloud?

Yes. It can run entirely on a server inside the plant, in the cloud, or split between the two. The choice usually comes down to your IT/OT policy and what the plant network can carry.

How long does the first implementation take?

It depends on how ready the data points are and how many machines are involved. We normally start with a pilot on one line or a few critical assets to prove out data quality and the workflow, then widen once it holds up on the floor.

Further reading

Plan Condition-Based Maintenance for Your Assets

Our engineers look at your machines and the data you already have, then scope a pilot that is realistic to run.

  • No obligation to buy
  • A reply within business hours
  • WhatsApp works too