Vibration sensors on 74 worst actors cut unplanned line stops by a third
Multi-plant manufacturing group, GCC. A manufacturing group running plants in two GCC countries moved from reactive firefighting to condition-based maintenance on its critical rotating equipment. Sensor-triggered work orders and honest downtime capture changed both the numbers and the weekly conversation.
Illustrative targets — customer-validated figures pending publication.
The situation
The group's plants ran a maintenance operation that looked organized on paper — an aging CMMS, OEM-derived PM schedules — but lived reactively. Over half of maintenance hours went to breakdowns. Downtime was logged manually at end of shift, so short stops vanished from the record and the downtime Pareto pointed at the wrong problems. The PM program consumed hours without visibly moving reliability.
Two production lines dominated the loss ledger, driven by a familiar cast: conveyor gearboxes, filler motors, compressors, and pumps that failed without warning. Each event triggered expedited parts orders and overtime, and the planning function spent its weeks rescheduling around emergencies rather than planning work.
What changed
The group replaced its legacy CMMS with OpsSense and, in the same program, instrumented its 74 worst-actor assets with Vibra vibration-and-temperature sensors connected through OpsSense Edge gateways. Run-state capture from the gateways replaced manual downtime logs: every stop is now timestamped automatically, and operators supply a reason code only for stops above five minutes, from a twelve-item picker at the line.
The Reliability Agent monitors the vibration and temperature baselines and raises inspection work orders — with the trend chart attached — when signatures drift. Every alert outcome is recorded as confirmed, false, or watch, and thresholds are tuned quarterly against that log. In parallel, a PM optimization pass retired roughly a fifth of inherited calendar tasks that had never produced a finding, freeing hours that were redirected into the condition program.
The result
In the first year the program caught eleven developing failures — bearing wear, misalignment, a failing compressor valve — all repaired as planned work. Unplanned downtime on the instrumented lines fell 34 percent, and the availability factor of OEE on the two pilot plants rose six points against the honest, sensor-measured baseline. Planned work now exceeds reactive work in hours for the first time in the group's records.
The second-order effects surprised leadership more. Automatic downtime capture rewrote the loss Pareto — minor stops, invisible in the old logs, turned out to be the largest availability loss on one line and were engineered out with two small fixes. And the verified-findings log turned budget season into a short conversation: the expansion to the remaining plants was approved on the program's own recorded numbers.
“The first bearing we changed on a planned stop instead of a breakdown paid for a lot of sensors. But the bigger change is the morning meeting — we argue about data now, not about whose fault it was.”
See your operations run themselves.
A 30-minute walkthrough with an operations engineer. Your assets, your workflows, your questions.