PlatformMaintenance
Fix — maintenance & tooling
Maintenance software that reports, routes, fixes, confirms.
Most breakdowns lose more time before anyone starts work than during the repair, and maintenance software usually only measures the repair. SmartFactory takes the reporting, finding, chasing and confirming out of the middle as well, and gives the maintenance manager the numbers to prove where the time went.
Reporting a fault
Under a minute, from the machine, by whoever found it.
If reporting a fault is harder than shouting down the line, people shout down the line. So the whole thing happens on a shop-floor tablet, standing at the press — inside the same system that already knows the machine has stopped.
- 1 Scan QR label on the press
- 2 PIN Operator identifies
- 3 Category Mechanical / electrical / tooling
- 4 Severity Machine down?
- 5 Photo Optional, from the tablet
- 6 Submitted WO-2026-0042 raised
- Open
- Assigned
- In progress
- Resolved
- Operator confirms
- Closed
Send it back instead and it bounces to In progress with the engineer emailed. A machine-down fault opens its own downtime event, timestamped from when the press actually stopped.
- Scan the QR label on the machine, enter a PIN, pick a category and a severity, add a photo — submitted. No account to create, no ticket system to learn.
- A machine-down fault opens a downtime event on its own, timestamped from when the press actually stopped, and rings that machine red on the status board immediately.
- The photo comes off the tablet’s camera at a sensible size, so a 6 MB snap does not sit in a queue behind a factory’s worth of Wi-Fi.
- Faults can be raised against a machine or a tool — a mould with a damaged cavity is not the press’s problem, and should not be recorded as one.
- The same fault can be raised from the web by anyone with an account — and where your policy allows phones on the floor, the same QR and the same flow work from one, with nothing to install.
Routing & escalation
The right team gets it, and the clock starts whether anyone is watching or not.
- Separate queues for press maintenance and the toolroom, with routing rules on category, machine, location and severity — because a hydraulic leak and a worn cavity are not the same team’s morning.
- Rules come with a “test this rule” explainer that shows you which team a given fault would land with, before you rely on it at 3am.
- An escalation ladder that actually walks: for a critical fault, the team at once, the lead at 5 minutes, the manager at 15, the administrators at 30 — and gentler ladders for lower priorities.
- Escalation notices are de-duplicated, so one unacknowledged fault does not generate the same email eleven times and train everyone to ignore it.
- Work can be put on hold with a stated reason — awaiting parts, awaiting a contractor, awaiting access, awaiting a decision — and hold time is subtracted from repair time rather than blamed on the engineer.
Closing the loop
It isn’t fixed because an engineer says so. It’s fixed because the operator says so.
This is the difference between a maintenance log and a maintenance system: the person who reported the problem is the person who confirms it has gone away.
- The engineer resolves the work order with a mandatory note — no one-word closures — and the machine moves to “awaiting confirmation” rather than straight back to green.
- The operator either confirms it fixed, or sends it back: it bounces to In Progress and the engineer is emailed. A repair that did not hold is visible instead of being re-reported as a new fault next week.
- Because the loop is closed, the metrics come for free: time to acknowledge, and repair MTTR measured from acknowledgement to resolution, less any time on hold.
- Repair MTTR sits next to downtime MTTR and is never quietly reconciled with it — one is how long the machine was down, the other is how long the work took, and the gap between them is usually the interesting number.
- Median and 90th-percentile are reported alongside the mean, by machine, tool, team, category and priority — so one catastrophic Friday does not hide a fortnight of small ones.
Tools & moulds
Moulds service themselves on shots, not on a calendar.
Maintenance software ships calendar-based preventive schedules because a calendar is all most systems have to work with. This one is already counting every cycle, so the schedule can follow usage instead.
T-0389 · Housing Rear 4-cav
- IMM-03 12 Jun – 04 Jul 38,420 shots
- IMM-07 05 Jul – today 21,940 shots
- Tools are first-class assets with their own history, status and documents — in service, in repair, quarantined or retired — not a text field on a machine record.
- Cycles accumulate per mount and per lifetime, and follow the tool across machines: move a mould from IMM-03 to IMM-07 and its count goes with it. Why that matters on a moulding floor, injection or compression.
- Usage-based preventive schedules raise their own work order at the threshold you set — “this mould is 4,000 shots from its 250,000 service” — with the due-soon warning arriving in time to plan it rather than discover it.
- Service work orders carry a mandatory checklist, including measured values where you need them recorded, so the completed job is evidence rather than a tick.
- Machines can carry usage-based and time-based schedules side by side: run hours for the press, shot count for the tool, months for the things that genuinely are seasonal.
Cost of repair
Every repair closes with a cost, so the replace-or-repair argument has numbers in it.
- Parts are booked against the work order with a price snapshot taken at the time of use — so last year’s repair still shows last year’s price after the supplier puts theirs up.
- Labour is booked from time recorded against the job at your own rate, and the closing cost card sums the two: £62 in parts plus £48 of labour, on the work order, permanently.
- Cost rolls up by machine and by tool, which is how a press that has quietly consumed four thousand pounds of parts this year becomes visible.
- Optional stock control, off by default: hold minimum levels and get a low-stock alert, or leave it off entirely.
- If your inventory system should own stock, it can — every part used fires a webhook, so the system of record stays the system of record.
Work that raises itself
The work order that exists before anyone has noticed the problem.
- Alert rules can raise work orders directly: a machine stopped for twenty minutes becomes a routed, escalating work order without anybody touching anything.
- Automatically raised orders are de-duplicated against anything already open on that machine, so one bad night does not produce forty tickets to close in the morning.
- Everything raised this way is tagged with where it came from — station, web, preventive schedule, alert or API — so you can audit how much of your maintenance load is finding you rather than being found.
- Maintenance sits on the same data as production: the downtime event, the work order and the cost per part on the affected job all point at the same shift.
SmartFactory covers production machines and tooling. It is not a facilities or general asset CMMS, and it does not claim to predict failures — preventive schedules are driven by measured usage and elapsed time, which is what your data can actually support.
See your factory live in thirty minutes.
Book a demo and we’ll walk you through a running factory — or register for a free trial and get a simulated factory of your own to play with, the same day. No hardware, no card, no IT project.