CMMS for Education and Campus Facilities

Most CMMS demonstrations are built for a factory or a hospital. The demo asset is a pump, the demo failure is a bearing, the demo constraint is production downtime. None of that maps cleanly onto an education estate. A school does not have production downtime, it has a thousand children arriving at 8.30am. A college schedules around a timetable that cannot move. A university has a mixed estate accumulated over a century, devolved budgets by faculty, and a six week summer window in which most of the disruptive work of the entire year must be executed. Selecting school maintenance software means judging products against those constraints rather than the ones in the demo script.

The message up front: in education the academic calendar is not a scheduling preference, it is the dominant constraint on the entire maintenance plan, and a CMMS that cannot model term time against vacation shutdown will fight you every year. Judge products first on calendar-aware planning, second on request handling at student and staff volume, third on contractor vetting and safeguarding controls, and only then on the feature grid. The harder problem is not the software at all: it is a deferred maintenance backlog that compounds faster than the budget grows, and the software's real job is to make that case visible enough to fund.

1. Why education estates break the standard CMMS assumptions

Before looking at products, it is worth being precise about what makes this sector distinctive, because each difference translates into a selection criterion later.

• Occupancy is rigid and non-negotiable. In an office you relocate a team for a morning. In a school you cannot relocate a class of thirty into a corridor while a radiator is replaced.
• Children and vulnerable adults are present. That changes contractor management from a procurement question into a safeguarding question, with vetting, escorting and access controls that industrial CMMS products rarely address natively.
• The estate is mixed to an extreme degree. A single campus can hold a listed nineteenth century building, a 1960s teaching block, a new laboratory wing, a sports centre with a pool, a catering kitchen and several halls of residence. Each is a different typology with a different maintenance regime.
• Request volume comes from amateurs, in bursts. Thousands of untrained students and staff raise requests, the volume spikes hard in the first fortnight of term, and requests arrive as "the room is cold" rather than as an asset reference.
• Budgets are devolved and politically held. Faculties, departments and academies hold their own money, so the CMMS has to answer "who pays for this" as reliably as "who fixes this".
• Funding is chronically tight and capital is bid for, not planned. Most education estates carry a known, quantified backlog they cannot afford to clear, which makes condition data and prioritisation the most valuable output the system produces.

If you are new to the category itself and want the generic groundwork before the sector specifics, start with the complete buyer's introduction to CMMS and the wider CMMS for facilities management guide. This article assumes you already know what a work order is and are trying to work out which product survives contact with a campus.

2. The academic calendar as the master schedule

This is the single point I would put ahead of every other requirement. In industry the maintenance plan is shaped by duty cycles and production windows. In education it is shaped by a calendar published a year or more in advance and completely immovable. Term dates, examination periods, open days, graduation, enrolment week and vacation shutdowns are fixed points, and the annual plan is built in the gaps between them.

The practical consequence is that the year divides into work classes rather than a smooth flow of planned maintenance. Nearly all intrusive, noisy, dusty or service-interrupting work has to be executed in a small number of vacation weeks, and the rest of the year is reserved for non-disruptive planned work and reactive response. A CMMS that spreads preventive maintenance evenly across twelve months, because that is what levelling the workload looks like in a factory, is actively working against the plan.

When I assess a product for an education client, this table becomes the test script. Three behaviours matter. Can the system hold named calendar periods and blackout windows as first-class objects rather than as a note in a description field. Can a PM task be assigned to a window class rather than a date, so that "annual fume cupboard testing, summer window" schedules itself into the right eight weeks each year without a planner rebuilding it. And when a request arrives during a blackout, does it route correctly rather than dropping into a queue where it is invisible until August.

Underneath this sits ordinary good PM discipline, which does not change because the sector does. The interval logic, the task content and the compliance backbone are the same as anywhere else, covered in the complete preventive maintenance guide and illustrated across sectors in preventive maintenance examples. What changes in education is when the work is permitted to happen.

3. Safeguarding and contractor vetting when children are present

This is the requirement most likely to be missing from a CMMS shortlisted on maintenance merit alone. Where children or vulnerable adults are on site, every person who enters must be accounted for, and in most jurisdictions background-checked or escorted at all times. In the United Kingdom that means enhanced disclosure through the Disclosure and Barring Service ; elsewhere the equivalent is a national police clearance or a sector-specific vetting regime. The obligation is on the institution, not the contractor, and it is enforced by inspection.

Translated into software requirements, the ones I insist on:

• Person-level compliance records, not company-level. It is not enough to record that a contracting firm holds a policy. You need vetting status, expiry date and evidence for each individual engineer who may attend.
• Hard blocks on assignment. An engineer whose clearance has lapsed should not be assignable to a work order in an occupied area. A warning is not a control. If the system can only warn, the control lives outside it, which means it will eventually be missed.
• Escort requirement as a property of the location. Areas where unescorted access is prohibited carry that flag, so the escort is planned rather than improvised at the gate.
• Sign in and sign out tied to the work order. A register you can reconstruct after the fact: who was on site, where, when, under whose supervision. This is what an inspection asks for.
• Insurance, method statement and risk assessment currency. The same expiry logic applied to documents, checked at assignment rather than at audit.

Very few CMMS products ship all of this. Some do person-level compliance well, some have location access flags, few enforce hard blocks. Where a gap exists, decide deliberately whether it is closed by configuration, by a separate contractor-management system integrated to the CMMS, or by a manual process with a named owner. What does not work is assuming the CMMS covers it because the word "compliance" appeared in the proposal.

4. The mixed estate: modelling a century of buildings in one asset register

The asset register on an education estate has to absorb an unusual spread of building types, and the structure you choose early determines whether reporting is useful later. A listed Victorian block with sash windows and a converted plant room sits next to a laboratory wing with fume extraction and gas services, next to a sports centre with a pool plant room, next to catering kitchens, next to halls of residence that are functionally small apartment blocks.

The register has to answer three questions at once: what is it, where is it, and who owns the budget for it. Get the hierarchy right and each is a filter. Get it wrong and every report becomes a manual exercise.

Two structural decisions matter more than the rest. Keep the location hierarchy deep enough to reach a room or a bedroom, because that is the granularity at which requests and turnover works actually happen. And separate the location hierarchy from the funding hierarchy, so a laboratory can sit physically in a science building and financially under a faculty without one distorting the other. For the underlying criticality method that decides which of these assets deserves the closest attention, the asset criticality classification guide applies unchanged; the life safety regime is covered in lift, fire and life safety system PM.

5. Request volume and the helpdesk model

Education generates request volume that is unusual in both quantity and quality. Thousands of potential requesters, no training, no asset vocabulary, and a strong seasonal spike. The pattern that works is a proper helpdesk: a single front door for requests, a triage step that turns a description into a work order, and status visibility back to the requester so the same issue is not reported eleven times.

What I would specify:

• Request by place, not by asset. A student reports a room number or scans a QR code on a door. Expecting a requester to identify an asset tag is a design error that produces unusable data.
• Institutional single sign-on. If students have to remember another password, they will email the estates office instead and your intake data collapses.
• Duplicate detection at the point of entry. When someone reports a fault in a location where an open work order already exists, show them that instead of creating a second ticket. In term time this one feature removes a meaningful share of triage effort.
• A short, fixed request taxonomy. Eight to twelve categories in plain language: too hot, too cold, light out, water leak, broken furniture, door or lock, IT or AV, cleaning, outside area. Long dropdowns get random answers.
• Automatic priority from category and location. A water leak in a laboratory is not the same priority as a broken chair in a common room, and the requester should not be deciding.
• Status notification without a login. Acknowledged, scheduled, done. Most perceived service failure in education estates is silence rather than slowness.

The general portal principles are in maintenance request and work request portals. The education-specific adjustment is that your user base turns over by a quarter or more every year, so the portal has to be learnable in thirty seconds by someone who will never be trained on it.

6. Halls of residence and the turnover window

Halls of residence are the part of the estate least like the rest of it. Functionally they are residential property: en-suite bathrooms, kitchens, domestic heating, white goods, door access, and occupants who live there rather than visit. That brings a residential compliance regime, in particular water hygiene and legionella control on systems that stand idle through vacations, fire door inspection, and gas or heat interface unit servicing per bedroom or flat.

The distinctive workflow is the annual turnover. At the end of the academic year several hundred or several thousand rooms empty within days, and a fixed number of weeks later they must all be ready for arrivals who have already paid. Turnover is not maintenance in the normal sense, it is a repeating campaign of inspection, repair, redecoration and cleaning against a hard deadline, and the software either supports it or gets abandoned in favour of a spreadsheet.

The capability that makes this workable is bulk work generation against a location set with per-room progress tracking: create the inspection for every bedroom in a block in one action, capture findings on a mobile device room by room, and let the findings generate repair work automatically. Without it, someone spends a week raising work orders by hand in July, which is exactly the week they should be sequencing the trades. Ask a vendor to show one thousand rooms of turnover created, inspected and reported on, and count the clicks a single room costs. Multiply it out honestly.

Also specify a condition record per room that persists between years. Knowing that a block's kitchens are three years past their expected refresh is a capital argument. Knowing only that they were repainted each August is not.

7. Laboratories, specialist equipment and research continuity

Laboratories concentrate the highest-consequence assets on most campuses, and they change the maintenance conversation in three ways.

First, a large share of lab work is statutory testing with certification rather than maintenance by condition: fume cupboard and local exhaust ventilation testing, autoclave pressure vessel examination, gas detection calibration, safety shower and eyewash testing, radiation and biosafety controls. The CMMS requirement is certificate lifecycle management: what was tested, by whom, to what standard, when it expires, and where the certificate lives. The UK Health and Safety Executive sets the expectations for most of these regimes in Britain, and comparable national bodies do elsewhere.

Second, research continuity creates criticality that has nothing to do with replacement cost. An ultra-low temperature freezer worth a modest sum can hold irreplaceable samples representing years of funded work. Its criticality derives from its contents, not its asset value: alarm monitoring, tested backup power, a documented failure response, spare capacity identified in advance. A criticality model that ranks by capital value will rank this asset far too low.

Third, control of access. Isolating services in a laboratory is not routine, and permit to work with academic sign-off is normal practice. The system should hold a work order in a state requiring a named responsible person's approval before release, and record that approval.

8. Statutory compliance across a large, dispersed estate

A multi-academy trust with thirty schools, or a university with two hundred buildings, faces the same compliance obligations as any other landlord but across a far larger and more fragmented footprint. The failure mode is rarely that a test was refused. It is that a test in one building, on one asset, was simply never scheduled, and nobody noticed until an inspection or an incident.

What prevents that is not more diligence, it is structure:

• A compliance asset register that is explicitly complete. Every asset subject to a statutory regime is flagged, and the absence of a scheduled test on a flagged asset is itself an exception report. Completeness has to be provable, not assumed.
• One compliance dashboard across all sites. Site-by-site reporting hides the gaps. Answer in one view: which statutory tests across the whole estate are overdue or expiring in the next ninety days.
• Evidence stored against the asset. Certificates attached to the asset and the work order, not filed in a shared drive by date. When an inspector asks for the last three fire alarm tests in a named building, that should be a two minute answer.
• Remedial actions tracked from findings. A test that identifies a defect creates a linked corrective work order with its own due date. Unclosed remedials are the most common compliance finding I see.
• Consistent standards across devolved sites. Where individual schools or faculties manage their own contractors, task content and frequency should still come from a central template rather than local habit.

Where the estate is genuinely multi-site, the data architecture question, shared templates and standards centrally with local execution and local reporting, is the same one covered in multi-site CAFM architecture. For a trust or a college group, that architecture decision is made once and lived with for a decade.

9. Devolved budgets, recharge and who pays

In most institutions the estates function does not hold all of the money. Faculties, departments, schools within a trust, and commercial operations such as conferencing or sports hire each hold budget, and a significant volume of work is either funded by them or recharged to them. Getting this wrong does not break maintenance, it breaks the relationship, and the estates team spends its year arguing about invoices instead of planning work.

Every work order needs attribution to a cost owner at the point of raising, using a defensible rule. The rule set I would agree in writing before configuration:

• Core estate maintenance, fabric, central plant, statutory compliance, life safety: funded centrally, never recharged. This has to be large and clearly defined or every conversation becomes a negotiation.
• Departmental equipment, research and teaching equipment owned by a faculty: recharged, with the department able to see committed spend against its own budget in real time.
• Alterations and improvements, anything that changes the space rather than maintaining it: recharged or project-funded, with an approval threshold above which a formal process applies.
• Damage and misuse: recharged where the responsible party is identifiable, which in residences means a clear evidence trail from the turnover inspection.
• Commercial and external hire: charged at a rate that genuinely recovers cost, which requires labour and materials captured accurately on the work order rather than estimated later.

Two software capabilities follow. Time and material capture has to be good enough to bill from, a higher bar than good enough to report from, and captured by technicians in the field without friction. And departmental users need read access to their own committed and actual spend, because disputes are almost always about surprise rather than amount.

10. Grounds, externals and the seasonal calendar

Grounds are frequently left out of the CMMS because they sit under a separate contract, and then reappear as a liability problem. External areas carry real risk: footpath trip hazards, car park lighting, tree condition, drainage, boundary fencing, play and sports equipment. Much of that carries a documented inspection frequency, and a claim will ask for the record.

The practical points. Externals need zone-based locations rather than buildings, because a path or a car park is not inside anything. Inspection-only work types matter, since most grounds compliance is a recorded inspection rather than an intervention. And the seasonal grounds calendar runs on weather and growing seasons, so it is almost the inverse of the building calendar: grounds work loads up in term time, precisely when work in occupied teaching spaces cannot happen, which is useful for labour levelling if you plan it deliberately. Where grounds are contracted out, at minimum pull the inspection records into the same system so the estate has one defensible record of external condition.

11. The honest section: chronic underfunding and the backlog that compounds

Everything above is a software selection exercise. This section is the reason the selection matters, and it is the part vendors do not want in the room.

Education estates are, across most countries, chronically underfunded for lifecycle replacement. Buildings reach the end of their design life and are not replaced. Roofs, windows, heating systems and electrical distribution are patched past the point where patching is economic. The result is a deferred maintenance backlog: a quantified list of work that is known, needed, costed, and unfunded. National audit and estates bodies publish backlog figures for their school and university sectors, and they are consistently large and consistently growing.

A backlog compounds in three ways. It gets more expensive, because a roof repair postponed becomes a roof replacement plus water damage to the ceilings, finishes and equipment below. It consumes revenue budget, because failing assets generate reactive callouts paid from the same pot that would have funded the planned replacement, so the more backlog you carry the less capacity you have to reduce it. And it displaces planned work in the vacation windows, because emergency repairs claim the scarce summer weeks allocated to lifecycle replacement.

That third mechanism is specific to education and it is the one most often missed. In a sector with fifty-two available working weeks, backlog steals money. In a sector with eight usable weeks, backlog steals the window, which is scarcer than the money. An institution can find emergency funding far more easily than it can find another summer.

12. Using condition surveys to make the capital case

The mechanism that turns a backlog into funding is the condition survey, and the discipline that makes it work is keeping it live in the CMMS rather than letting it sit as a consultant's PDF that ages out.

The approach I would recommend:

• Survey by building element, on a consistent grading scale. Roof, walls, windows, internal finishes, mechanical services, electrical services, externals, each graded for condition and urgency. A simple published scale applied consistently beats a sophisticated one applied inconsistently.
• Grade condition and priority separately. A poor-condition element in a rarely used store is a different investment case from a fair-condition element housing your main lecture theatres. Two axes, condition and consequence, produce a defensible ranking. This is asset criticality classification applied to building fabric.
• Cost every line. An uncosted backlog is an opinion. A costed backlog, even with a stated accuracy range, is a budget submission. Recost annually against actual tender prices rather than letting original estimates drift.
• Keep it live in the CMMS. Every completed major repair should close out the corresponding survey line, and every significant reactive failure should be able to create or upgrade one. A survey refreshed on a five year cycle is out of date for four of those years.
• Link reactive spend back to survey lines. Showing that a named building consumed emergency revenue spend last year against a known unfunded backlog item reframes the capital request as cost avoidance rather than a wish.
• Report the trend, not just the total. Whether the backlog grew or shrank this year, and what it would take to hold it flat, is the number a finance committee can act on. A single large total invites despair; a trend invites a decision.

The reactive spend argument depends on tracking that most institutions do poorly, so it is worth building deliberately. The method is in maintenance backlog and downtime tracking, and the reporting frame that carries it to a governing body is in the facility management KPI framework. For an estates-standards reference point when building the survey scale, the Royal Institution of Chartered Surveyors publishes the building survey and condition grading conventions most UK and international consultants work to.

The idea to walk away with

Choosing school maintenance software is not a search for the longest feature list. It is a search for the product that respects four constraints the rest of the CMMS market does not design for: a calendar that cannot move, children on site who change who may enter, an estate too mixed for one template, and request volume generated by untrained people in seasonal bursts. A product that handles those four will handle ordinary maintenance work comfortably. A product that handles the ordinary work beautifully and fails on the four will be worked around within a year.

And the deeper point: the system's most valuable output in this sector is not efficiency, it is evidence. A quantified, prioritised, costed, defensible picture of what the estate needs and what is not being funded, produced consistently enough that a governing body can act on it. Institutions that get that right do not escape underfunding, but they direct the money they do get at the highest-risk work and can prove they did. The rest spend the same money on whatever broke most recently.

Final thoughts

If you are a school or trust starting from spreadsheets, do not start with a long selection exercise. Build the asset and location hierarchy down to room level, get the statutory compliance register provably complete, and open a simple request channel. Those three deliver most of the available value and are the foundation everything else needs. Sophistication can wait; a missing fire alarm test cannot.

If you are a university or a college group, the two decisions that will still matter in ten years are the multi-site data architecture, meaning what is standardised centrally and what is devolved, and whether the condition survey lives inside the system as maintained data or outside it as a document. Both are easy to get wrong early and expensive to change later. For the higher-education-specific architecture and governance questions in more depth, including multi-campus structure, procurement, space management and research facilities, the companion piece is CAFM for universities, which goes considerably further on the university estate than this broader sector guide does.

Wherever you sit on that range, run the calendar test first. Put your own published term dates in front of a vendor and ask them to plan a year against them. How they respond tells you within an hour whether they have done education before or are about to learn on your estate.

Related reading: CAFM for universities, CMMS for facilities management, What is a CMMS: a complete buyer's introduction, Maintenance request and work request portals, Multi-site CAFM architecture, Maintenance backlog and downtime tracking.