Spot equipment drifting from its normal pattern, with the signals that moved.
Fixed alarm limits catch a machine once it crosses a line. Slow drift below that line sits unseen in your historian. A check of each asset against its own normal pattern is designed to flag that drift to the engineer who owns the asset, with the readings that changed.
What is equipment anomaly detection?
Equipment anomaly detection compares each machine’s sensor readings with its own normal pattern and flags readings that break from it, so an engineer can look early. Agentra’s planned Predictive Maintenance is designed to check your data first, flag drift with the signals that moved and route each flag to the asset’s owner. It never controls equipment.
Why do fixed alarm limits miss early changes?
They wait for a line to be crossed. Switch views to see where drift goes unnoticed.
- 1Call rings outYour tech is on a job0:30
- 2Voicemail leftCaller1:00
- 3Call back, no answerOffice1:30
- 4Call back againOffice2:00
- 5Check the calendarOffice1:30
- 6Agree a windowOffice and caller1:30
- 7Enter the bookingOffice1:00
What should raise a flag?
You decide with your engineers. Switch rules on and off to see what a flag would mean.
It checks the address is in your area, offers tomorrow onwards, leaves 30 minutes between visits, stops at 6 visits a day, and confirms the booking straight away.
What do you need before you start?
- Sensor or historian dataReadings such as vibration, temperature, pressure or current, tied to asset IDs, with enough history to show normal running.
- Maintenance historyPast work orders and findings for the same assets, so flags can be checked against what actually happened.
- An owner for each assetThe engineer who reviews its flags, and the lead a flag goes to if nobody responds in time.
- Your maintenance systemWhere work orders are created, always by a person. A connection is planned for a later release. {no claim row; Q33}
What could go wrong, and what happens instead?
- Not enough dataAn asset with gaps or no history is marked insufficient data and is not scored, rather than given a guess.
- Nuisance flagsEngineers dismiss them with a note. Thresholds are tuned only within limits your admin sets.
- A new normalAfter a rebuild or a new product run, the pattern changes. It is relearned only after the owner confirms.
- A safety-critical conditionIt is never the only alarm. Your existing safety systems and alarm limits stay in force.
Which measures should a pilot agree first?
Which Agentra products does this use?
Last updated
Questions, answered.
Not yet. Anomaly flags are planned in Predictive Maintenance. We are exploring it with maintenance and reliability teams before it is built, and each site would start with a data feasibility check.
Not as a promise. It is designed to flag behavior that departs from normal and show which signals moved. Forecasting comes later, and only where there is enough failure history.
Alarms trip when a reading crosses a fixed limit. Anomaly detection compares each asset with its own normal pattern, so it can flag a slow change while readings are still inside your limits. Your alarms stay exactly as they are.
No. Agentra sells no sensors. It is designed to use the readings you already collect, where they exist. If an asset has no data, it is a poor fit, and we will say so.
No. It has no control connection to equipment and takes no safety action. Work orders are created by a person in your own maintenance system.
Predictive Maintenance starts at US$349 a facility a month for up to 10 assets, with onboarding per facility for the data connection and baseline. A six-week pilot is US$3,500, credited in full against your first invoices when you subscribe.
These terms apply to pilots and plans agreed in a written order form. Every solution is In discovery.
Bring us one process. We’ll show you how it would run.
A 30-minute call with a specialist: what it connects to, what a pilot involves, and an honest view on fit, including when it isn’t one.
- 30 minutesOne call, one process
- No commitmentNo sales script
- 4 business hoursTypical reply time · Mon to Fri, 7:00 to 22:00 IST
- 1You get a short agenda by email.
- 2We look at one process with you and how it would run.
- 3You get an honest view on fit, and what a pilot would involve.