About the project

The Predictive Maintenance System moves the plant from fixed interval servicing to condition based servicing. Wireless sensors combining vibration and temperature are mounted on the drive end and non drive end bearings of critical motors, reporting continuously through an industrial gateway and 4G router to a condition monitoring platform that plant staff reach from a SIMATIC HMI panel on the floor or from a browser anywhere on the network. Our scope covered selection of the machines to monitor, sensor mounting and commissioning, gateway and network setup, baseline measurement, alarm threshold configuration, and training the maintenance team to read trend, spectrum and time waveform so that a warning can be traced to the component that is actually failing.

Benefits of Project

  • Developing faults detected weeks before they cause a breakdown
  • Maintenance planned into shutdown windows instead of emergencies
  • Bearings and couplings replaced on condition rather than on a fixed calendar
  • No cabling to the measurement point, so installation does not stop production
  • Machine health visible to the maintenance team and to management at the same time
  • Objective vibration and temperature history to support equipment replacement decisions

Technologies We Used

  • Wireless vibration and temperature sensors, one unit per measurement point
  • Triaxial measurement: X, Y and Z acceleration and velocity
  • Envelope measurement for early bearing fault detection
  • Industrial gateway in a stainless enclosure with external antennas
  • Teltonika RUT950 industrial 4G router for remote data transfer
  • SKF Insight IMS condition monitoring platform
  • SIMATIC HMI panel for plant floor access to the dashboard

Condition Monitoring Without New Cable

Each measurement point carries a single wireless sensor that combines vibration and temperature in one housing, mounted on the motor bearing at the drive end and non drive end. Readings travel to a gateway in a stainless enclosure with external antennas, and from there over an industrial 4G router to the monitoring platform. Because nothing has to be wired back to a panel, sensors can be added to a running machine and the system can be extended one point at a time.

From Health Status to Root Cause

The dashboard ranks every monitored asset as Good, Satisfactory, Unsatisfactory or Unacceptable, and separates alarms into Need Action and Need Attention so the team knows what to look at first. Behind that summary sits the full diagnostic picture: long term trend, FFT spectrum, time waveform and envelope for each axis, plus temperature and sensor battery level. That is the difference between knowing a machine is getting worse and knowing which bearing is failing.