Maintenance

Keeping plant running, and being able to prove that it did.

For maintenance planners, foremen, supervisors and plant managers.

Printable: Maintenance user guide (PDF) — the same guide laid out for a desk, a cab or a workshop wall.

What this module is for

Maintenance answers three questions that a spreadsheet answers badly: what is due, what is being worked on, and how much of the time the plant was actually available. Everything in the module exists to serve one of those three.

It sits on top of the assets you already keep in Trackapt. You do not re-enter your plant here — a pump already tracked in Trackapt (a Tracked Item in a Tracker Group) is the same pump the module schedules, raises jobs against and measures downtime for.

What it does not do

Knowing the edges of a module saves a great deal of searching for a screen that was never built.

• It does not hold your stores. Parts are issued on your store's own form, raised from the job so each issue is linked to it and counted there. The stock itself lives where it already lived, and the count is of issues — not of parts, and not of money.
• It does not pay anybody. Labour hours are captured on the job because they tell you what the work cost in time, not so that somebody can be paid from them.
• It does not decide your maintenance strategy. You tell it the interval; it tells you when that interval has run out.

Who can see what

The module inherits the access rules you already use. A member sees the assets in the tracker groups they are whitelisted on, and nothing else. Jobs, stoppages and schedules for an asset follow that same boundary.

WARNING

Access comes from being whitelisted on the tracker group, not from your clearance level and not from your department. A senior member who has not been added to a group will not see its assets here. If plant is missing from your screens, ask the group owner or the account holder to add you.

Setting it up

Three things to put in place, in this order. You can start using the module after the first two; the third is what turns it from a job list into a measurement.

1. Confirm the module is switched on

Maintenance is an optional module. When it is active on your account, Maintenance appears in the left sidebar with five entries beneath it. If you cannot see it, it has not been enabled — that is an account setting, not something you can turn on yourself.

2. Register your assets

Open Maintenance → Assets and press Register an asset. Choose the Equipment class and the Equipment in it, give it its meter and commissioning date, and set its criticality — low, medium or high. Criticality does not change any calculation; it orders the work so that when six things are overdue you can see which one matters.

An equipment class is a Tracker Group marked as Equipment, and the picker offers only those. The Equipment tick is set when a group is created and cannot be changed afterwards, so if the picker is empty, create a Tracker Group for the plant and tick Equipment as you do.

Where an asset has a meter — hours, kilometres, tonnes — make sure it is being read. The module uses the meter you already capture; it does not introduce a second one.

3. Write your services

A service is a recurring piece of work and the interval it recurs on. This is the part worth doing carefully, because everything the module tells you afterwards follows from it. See Writing a service.

TIP

You do not have to schedule everything before the module is useful. Raise jobs and record stoppages from day one; add services as you work out what the real intervals are. The availability figures start accumulating the moment you record your first stoppage.

The screens, one at a time

Maintenance

Sidebar: Maintenance

The landing screen, and the one to open first thing in the morning. It leads with Needs attention: every service that has fallen due or gone overdue, worst first.

Each row tells you:

Column	What it means
Asset	The machine the service belongs to.
Service	The name you gave the recurring work.
Standing	OK, Due, Overdue, or Unknown.
Remaining	How much interval is left — days, or meter units, whichever is running out first.
Measured from	Whether the countdown starts at the last completed service, or at commissioning because there has never been one.
Open jobs	How many work orders are already open on that asset, so you do not raise a second one for the same thing.

Services

Maintenance → Services

Where recurring work is defined. Each service carries a title, the tasks it involves, what it applies to, and its interval. The list shows the current standing of every one.

A service can apply either to one asset or to a whole equipment class. The second is the one that saves real time: "500-hour service, all generators" is written once and follows every generator in the class, including ones added later.

Work orders

Maintenance → Work orders

The job queue. Everything currently being worked on, plus everything waiting. Open any job to see its history, record what was done, log hours and parts, and move it to its next status.

Downtime

Maintenance → Downtime

Stoppages, which are the raw material of every availability figure the module produces. The screen is in two halves: what is down now, and the finished stoppages behind it.

A stoppage is an interval — it has a start and an end. That makes it different from everything else in Trackapt, which records values at points in time, and it is the reason this module can produce availability at all.

Assets

Maintenance → Assets

The register: every asset, its tag, its meter, its criticality and how many jobs are open against it. Open one to see its full history — every job, every stoppage, in order.

Availability

Maintenance → Availability

The measurement screen. Availability, MTBF, MTTR, total downtime, stoppage and job counts, hours booked and stores issues — over whatever window you choose. How the figures are worked out explains each.

INFO

This is the module's reporting screen. Reports in the sidebar is a different thing — the report composer.

Everyday tasks

Writing a service

1. Open Maintenance → Services and choose New schedule.
2. Give it a title somebody else would recognise — "250-hour engine service", not "Service A".
3. Choose what it applies to: one asset, or a whole equipment class. Only equipment classes are offered — a service is an interval on machines.
4. Set the interval. You may set a time interval, a meter interval, or both.
5. List the tasks it involves. These become the checklist on every job the service raises.
6. Save. The service appears in the list with its standing worked out immediately.

Setting both intervals is normal, and the stricter one wins

Give a service 250 hours and 6 months and it falls due at whichever arrives first. The screen tells you which one is driving the verdict, so a machine that has sat idle all quarter shows as due on time rather than on hours.

WARNING

A meter interval on an asset whose meter is not being read cannot be answered. The standing shows as Unknown rather than OK — the module will not treat a missing reading as a zero reading, because a service reported as fine when nobody knows is the more dangerous of the two errors.

Raising a job

1. From Needs attention, use Raise a job on the row. The asset, the service and its task list are filled in for you.
2. For unplanned work, use Raise a job on the Work orders screen instead and choose the asset yourself.
3. Set the origin — where the work came from. This is what lets the Availability screen later tell you how much of your effort was planned and how much was firefighting.
4. Set a priority, describe the fault or the work, and save.

Origin	Use it when
Schedule	The work came from a service falling due. Set for you when you raise from Needs attention.
Fault	Something broke.
Inspection	Somebody found it while looking at something else.
Request	Somebody asked for it.

Working a job through

Open the job and use Move it along. A job travels through these statuses:

Status	Meaning
Raised	Logged, not yet planned.
Planned	Scheduled, parts and people identified.
In progress	Being worked on now.
Held	Started but stopped — waiting for a part, a permit, a window.
Completed	Finished. Records the work done, the hours and the meter.
Cancelled	Not going to happen. Kept, not deleted, so the history stays honest.

Completing a job is what resets the service interval

Until a job is completed, the service it came from stays due — which is correct, because the work has not been done yet. This is also why the module never asks you to enter a "next service date": it works it out from the last completed job, every time you look.

Reopening a completed job

A job closed too early, or against the wrong machine, can be taken back. Open it and choose Reopen job; you are asked why, and the reason is kept in the job's history. You need the same level to reopen a job as to complete one.

The job goes back to the status it had just before it was completed — a job that was Held goes back to Held. Its completion date and meter reading are cleared, so the next service of its schedule is measured from the completion before this one. If a later service of the same schedule has been completed since, that later one still sets the countdown. Nothing is deleted: the history keeps what the completion recorded.

Issuing parts to a job

Parts are issued on your store's own form, raised from the job itself. On the job, under Stores issues, choose Raise stores issue. The form opens with the job named at the top; fill it in as the store always does. Saving it links it to the job, and when you leave the form you are taken back to the job.

• Only forms marked Stores issue are offered. Whoever builds your forms ticks that on the one your store uses, beside One entry per item, per period. A fuel log or an inspection is never offered, so it can never be counted as parts.
• An issue belongs to one job. It cannot be moved to another; if it went on the wrong job, delete it and raise it again from the right one. Deleting an issue takes it off its job.
• A draft is linked to the job straight away but not counted: nothing has left the store until the storeman files it.
• A completed job can still take an issue — parts are often booked out the morning after the card is closed. A cancelled job cannot.
• A job whose earlier issues went out on paper still takes new issues, but its list says it may be partial.

Recording a stoppage

1. Open Maintenance → Downtime and choose Record a stoppage.
2. Pick the asset, the time it started, and the reason.
3. Leave the end time empty while it is still down. It appears under Down now.
4. When it comes back, open the stoppage and set the time it ended.

WARNING

Leave a stoppage open and every availability figure that includes it is a lower bound on downtime — the machine is counted as down right up to the moment you look. The Availability screen says so when this is happening, but a figure exported while a stoppage is open will keep getting worse until somebody closes it.

TIP

Do not worry about two people logging the same breakdown. Overlapping stoppages are merged before anything is calculated, so the same minute cannot be counted twice.

How the figures are worked out

Every number in this module can be checked by hand: if you cannot reproduce a figure on paper, you cannot defend it in a meeting.

Due, overdue, and unknown

Nothing about due-ness is stored. It is worked out afresh every time you open a screen, from the last completed job on that asset. Correct a job recorded on the wrong day and the standing corrects itself immediately.

Standing	What it means
OK	Interval has not run out.
Due	Interval has run out; work is now expected.
Overdue	Interval ran out and remains undone.
Unknown	Cannot be answered — usually a meter interval on an asset with no reading. Not the same as OK.

Measured from tells you where the countdown started: the last completed service, or commissioning where there has never been one. An asset that has never been serviced still gets an honest answer.

Availability

Uptime as a share of the window you selected. Stoppages are first clipped to that window and any that overlap are merged, then the merged downtime is subtracted.

MTBF and MTTR

• MTBF — mean time between failures. Uptime divided by the number of stoppages.
• MTTR — mean time to repair. Downtime divided by the number of closed stoppages only.

INFO

Open stoppages are left out of MTTR rather than counted at their length so far. A repair still in progress has no duration yet, and including it would drag the average toward whatever time of day you happened to look.

Where the work went

The breakdown by origin. It is the figure to watch over months rather than days: work arriving mostly as Fault means the schedule is not catching things, and that is a planning problem rather than a maintenance one.

Stores issues counts the filed issues raised from each machine's jobs. It is a count of documents — one issue can be a split pin or a gearbox — not a quantity of parts and not a cost; stores forms carry no price. Drafts are not counted.

• not linked — no job behind the figure has a linked issue; their issues went out on paper. Nothing was recorded; it does not mean no parts were used.
• 3 issues · 2 earlier jobs partly or not linked — some jobs had issues go out on paper, which are missing here, so the 3 is a floor.
• A plain number is whole: every job behind it could have had every issue linked.

When something looks wrong

What you see	Why, and what to do
An asset is missing	You are not whitelisted on the Tracker Group holding its equipment class. Ask the group owner or the account holder to add you.
The Equipment class picker is empty, or a class is not offered	Only Tracker Groups marked as Equipment hold assets, and the tick is set at creation. Create a Tracker Group for the plant with Equipment ticked; an existing group cannot be changed.
A service says Unknown	It has a meter interval and the asset has no meter reading. Capture a reading; the standing resolves at once.
A service stays Due after the work	The job has not been Completed. Only completion resets the interval.
Availability looks too low	Check for a stoppage still open under Down now. An unclosed stoppage counts as downtime up to this moment.
Downtime is longer than the window	It cannot be — overlapping stoppages are merged first. If you are comparing against a hand-kept figure, the hand-kept one is probably double-counting an overlap.
Two identical jobs on one asset	Somebody raised a second job for work already queued. Cancel one; cancelled jobs are kept so the history still reads correctly.
A job was completed by mistake	Open it and choose Reopen job, saying why. The next service is measured from the completion before it again.
There is no Raise stores issue on a job	No form is marked Stores issue yet, or the job is cancelled. Ask whoever builds your forms to tick it on the store's issue form.
Stores issues read not linked or partly or not linked	Those jobs had issues go out on paper rather than linked to the job. It is not a zero.
A stores issue is on the wrong job	An issue cannot be moved. Delete it and raise it again from the right job.

Reference

Taking the data with you

Every screen that shows meaningful data offers the same set of actions in its header: a reading-mode preview, Excel, PDF, print and share. What you export is what the screen shows — filters and date ranges included — so narrow the view first, then export. On the two lists that come a page at a time, Assets and Work orders, that means the whole filtered list, every page of it, up to 5,000 rows; past that the screen says so, and the export's subtitle says which part it holds. See Taking data away.

Words this module uses

Term	Meaning
Asset	A machine you maintain. It is already in Trackapt as a Tracked Item; the module adds maintenance facts to it.
Equipment class	The kind of machine an asset is — generators, haul trucks, pumps. In Trackapt, a Tracker Group marked as Equipment.
Service	Recurring work plus the interval it recurs on.
Work order / job	One instance of work on one asset.
Origin	Where a job came from: schedule, fault, inspection or request.
Stoppage	An interval during which an asset was not available.
Criticality	How much an asset matters. Orders the work; changes no calculation.
Availability	Uptime as a share of a window.
MTBF / MTTR	Mean time between failures / mean time to repair.