Maintenance
Meters that count the hours and cycles a machine has actually run, plans that come due on a date, a reading or a condition, and requests raised by whoever found the problem.
Maintenance goes wrong in the gap between the machine and the paperwork. The hour meter is on the control, the schedule is in a binder, and the request is a note left on someone's desk, so a plan comes due against a number nobody has read in a month. Here the meter lives on the resource that runs the job, every reading is a dated record with its own delta, and a plan points at the meter it depends on. The same resource carries the routing operations production schedules against it, so the machine you maintain and the machine you load are one record rather than two lists that drift.
Meters On The Machine
A meter belongs to a resource and is either a counter or a gauge, reading in hours, cycles, pieces, miles or a unit you name. Current and lifetime readings are held separately, so a replacement resets the count without losing the history.
Readings With A Delta
Every reading records the value, the change since the last one, the running lifetime figure and the moment it was taken. Readings arrive by hand or from telemetry, and a reset is marked as a reset rather than looking like a wild swing.
Three Ways To Come Due
A plan triggers on a calendar interval, on a meter reading interval, or on a condition expression. That covers the annual inspection, the five hundred hour service and the check that only matters when something else is true.
Plans That Look Ahead
A plan carries a lead time and a next due date, and records what it last raised and at what reading. The lead time is what lets the work land on a schedule before the machine is already overdue.
Tasks With Limits
A plan holds an ordered task list. A task can require a measurement with an upper and lower limit in a named unit, so the check records a number against a tolerance instead of a tick, and a safety step is flagged as one.
Requests From The Floor
Anyone can raise a numbered request against a resource with a description, a priority from low to urgent, and a flag for whether the machine is down. It records who reported it and when, so the queue is not a verbal one.
Approval On The Record
A request is open, approved or rejected, and a rejection carries its reason and the person who made the call. That is the audit trail for work somebody decided not to do, which is the half usually missing.
If you are shopping for a maintenance management system, this is the part of WorkCell you are comparing: meters, triggered plans, task lists and a request queue against the asset.
Manufacturing ERPMaintenance here is not a separate database bolted beside production. The asset it maintains is the same resource that routings schedule against and that jobs are costed through.
The resources maintenance looks after are the ones work orders are scheduled against and labour is booked to.
EngineeringResources and resource groups are defined in engineering, alongside the routings that decide which machine runs what.
CMMS SoftwareWhat a maintenance management system is, what it covers, and how it sits next to the rest of a manufacturing stack.
MTBF CalculatorWork out mean time between failures and mean time to repair from your own operating hours and failure count.
What is the difference between a plan and a request?
A plan is work you decided on in advance, and it comes due on its own when a date arrives, a meter passes an interval, or a condition holds. A request is work somebody found. It is raised against the machine by the person who found it, with a priority and a note on whether the machine is down.
How do meter readings get into the system?
Either a person enters one or it arrives from telemetry, and the reading records which. Each one stores the value, the delta since the previous reading and the running lifetime total, so a plan that triggers on an interval has a dated series to measure against rather than one current number.
What happens to the meter when a machine is rebuilt?
Mark the reading as a reset. The current reading starts again while the lifetime figure keeps climbing, so a plan on a five hundred hour interval measures from the rebuild and the asset still shows everything it has ever run. Without that split a rebuild looks like a meter fault.
Why would a maintenance task need a measurement?
Because a tick tells you someone looked and a number tells you what they found. A task can require a measurement with an upper and lower limit in a named unit, so a belt tension or a clearance is recorded as a value against a tolerance and the trend is visible across services.
Can maintenance see what production has the machine doing?
Yes, because it is the same resource. The machine a meter counts is the machine routing operations are scheduled against, so the resource a plan is coming due on is the one carrying load. A down flag on a request is therefore about a machine the schedule already knows.
One database, every module.
Tell us what you run and how you track service on it today, and we will walk through meters, triggered plans and the request queue against your own machines.