CapabilitiesMachine health and predictive maintenance
Machine health and predictive maintenance
Investigate a change in machine condition, decide what to check and follow the inspection into maintenance. A useful finding gives the planner the affected component, supporting readings and a practical way to test the suspected cause.
Product documentation · Updated 15 November 2026
- Who uses it
- Reliability engineers, maintenance planners and technicians
#Detect and diagnose a deviation
Compare the reading against normal behaviour for the same asset and operating state. Inspect the baseline, trend, quality flags and other available measurements. ORCAS is the intended diagnosis and prognosis engine. Ultron presents the evidence, candidate causes and local verification step.
#Interpret a forecast
Identify the predicted event and the assumptions behind its date. In the simulated primary-air fan case, drive-end vibration was 3.35 mm/s RMS against a normal value of 2.15. With a 4.5 mm/s high alert and a rise of about 0.4 mm/s per week, the projected crossing was about 2.9 weeks away. The calculation estimates time to a vibration limit; equipment failure is a separate event.
#Request and verify an inspection
In the demo, the planner requested vibration analysis from the PM cell. REQ-013 retained the asset, suspected cause, priority and projected alarm, and moved to In progress when started. In a plant, a technician would enter the field readings and finding. Preventive, predictive and proactive maintenance views serve different planning needs.
#Outcome and verification
- What the team gets
- A diagnostic record and traceable inspection request. A verified field outcome is added when the technician completes the inspection.
- How to verify it
- Follow one finding into the maintenance queue. Check that the asset, source readings and suspected cause survive the handoff and that the technician can return a finding.