What you own is the easy question.
Enterprise asset management is usually sold as a bigger maintenance system. It isn't. The register is the easy half — the hard half is what each machine must do, how it fails, what you decided about that, whether the decision is still in force, and what it will cost to keep versus replace. Zenith holds all five as records with evidence, not as a spreadsheet somebody refreshes each spring.
ISO 55000 · ISO 14224 · RCM decision logic · Bilingual EN / FR · Multi-siteSix records, one machine
An asset register that holds identity and location answers one question. Zenith builds five more into the same record, so “what is this, why does it keep breaking, and what do we do about it?” is never scattered across a drawing set, a binder and the memory of whoever has been here longest.
Identity
Class, type, criticality, health and site scope — the controlled backbone every other module points at, and the one thing that must never be typed twice.
The engineering record
Specifications with a stated source — manufacturer, vendor, commissioning, site-verified, field-measured — so a nameplate rating and somebody's measurement are never confused.
How it fails
Functions with measurable performance standards, functional failures, modes, causes and effects — aligned with ISO 14224 and reusable, so a bearing failure defined once serves every machine of that class.
What you decided
A strategy per failure mode, with the rationale, who approved it, which version is in force and when it is next due for review.
What actually happened
Coded failure history kept strictly apart from the library of what can happen — and reliability figures computed from it that name their evidence.
What it costs to keep
Installed, commissioned, useful life, replacement cost, warranty — beside the planned and actual cost of the work, which is what a replace-or-repair conversation actually needs.
A strategy is a controlled document, not a preference
Most systems record the PM. Almost none record why that PM exists rather than a different one, who agreed to it, or when somebody last checked it was still the right answer. Five years on that gap is the whole problem: the task survives and the reasoning evaporates, and nobody dares change it because nobody knows what it was protecting against.
- Thirteen strategy types — condition-based, predictive, scheduled restoration, failure-finding, run-to-failure, redesign, operator care and more, reached by answering five questions in the order an engineer actually thinks about them.
- The path suggests; the engineer decides — you can override it, and the reasoning is recorded next to what was suggested, so the decision stays explicable years later.
- Publishing supersedes rather than overwrites — the version in force is named, and the one it replaced is still readable, because work was planned under the old text.
- Review dates are tracked and surfaced — strategies past their review date are listed, because that is the part of reliability-centred maintenance that always lapses first.
- Doing nothing is never offered for a hidden or safety-related failure — run-to-failure is reachable only when the consequence is purely economic.
A number nobody can trace is a rumour
The temptation in reliability reporting is to fill every box. Zenith refuses: a metric is either computed — with the evidence it used named — or it is empty, with the reason. “PM compliance 0%” on a machine with no PM data is worse than showing nothing, because somebody will act on it.
- Fourteen metrics over a stated window — MTBF, MTTR, availability, downtime, failure rate, repeat failures, emergency ratio, planned-versus-reactive, cost.
- MTBF refuses below two failures — one failure has no “between”, and a machine with one failure is not infinitely reliable; it is one failure old.
- The basis travels with the figure — what was counted and what was excluded, by name, so a challenge lands on the data rather than on the person presenting it.
- A failure is a finished corrective job carrying failure coding — dated when the machine went down, not when the paperwork closed.
- Reproducible — every calculation takes the date as an input, so the same question asked twice gives the same answer.
Repair or replace, with the workings shown
The ledger already knows what a machine has cost to keep — the jobs, the hours, the parts, the downtime. What it cannot know is what a new one costs to buy, so you supply that and the arithmetic answers with its rule printed and every job it counted cited by code. A recommendation nobody can audit is an opinion with a currency symbol on it.
- The rule is printed, not implied — the threshold it applied is on the screen beside the answer, so the same question asked by finance and by engineering produces one number.
- Every job it counted is listed — as links, so “which repairs?” is a click rather than a request to somebody.
- Unpriced work is declared — hours booked with no rate against them are named beside the total rather than quietly counted as free.
- Lifecycle figures live on the machine — installed, commissioned, useful life, replacement cost and warranty, rather than in a parallel workbook that ages differently.
- It does not decide for you — it assembles the evidence for a decision a person makes and signs, which is the only kind that survives a budget review.
Your scheme, not a number we invented
Risk scoring is your organization's method — its bands, its thresholds, its consequence categories. Zenith stores the resulting score against the specific machine-and-mode pair, keeps assessments append-only, and flags a score made under an older scheme rather than silently comparing it to a newer one.
- Assessments are append-only — a reassessment is a new record, so how your view of a machine changed over time is readable rather than overwritten.
- An older scheme is flagged, not converted — because quietly restating a 2021 score in 2025 terms is how a risk register stops meaning anything.
- Criticality is separate from risk — what the machine is worth to the operation, and what a particular failure of it would cost, are different questions with different owners.
- Worst-of health rolls up the hierarchy — a branch is coloured by the worst machine under it, because green over a critical asset is actively misleading.
Five ways to walk the same estate
Facilities teams think in buildings. Reliability engineers think in systems. Planners think in classes. Finance thinks in sites. All five trees are rebuilt from the one register every time it changes, so none of them can drift out of step with the others — or with the truth.
Enterprise, site, system, class, and parent-and-child
The same assets, walked whichever way the question demands: the whole estate, the location registry, what hangs off this machine, a functional system that spans three buildings, or every pump you own for the day the question is “show me all of them”.
- An enterprise figure that is the sum of its sites — one row per site with open work, overdue, in progress and PM compliance, each computed by the same folds the site's own page uses.
- Bad actors ranked on hours lost — with every work order the ranking counted attached, because a bad-actor list nobody can audit is a rumour with a sort order.
- Fourteen typed relationships — feeds, powers, controls, backs up — held as dated records, so “when did this change?” has an answer.
- Site scope enforced in the query — not by hiding a menu, so a portfolio view and a site view cannot disagree about what exists.
What an EAM buyer asks that we do not do
EAM is the category where the acronym covers the widest ground, so the boundary matters more than usual. Four things are commonly expected under that heading and are not here — stated before you score them rather than after.
The asset record, not the general ledger
Zenith holds what an asset is, what it does, how it fails, what you decided and what it has cost you to keep. What it deliberately does not hold is the finance department's version of the same machine — and pretending otherwise is how an EAM implementation becomes a two-year reconciliation project.
- Depreciation belongs to your financial system — book value and maintenance cost answer to different people under different rules, and one record trying to be both satisfies neither.
- We hold the inputs to a capital plan, not the plan — condition, criticality, useful life, replacement cost and failure history, each traceable to the jobs behind it.
- No curve-fitting we cannot explain — the only projection in the product is a straight line that publishes its fit and its point count, because a model nobody can argue with does not change a decision.
- The audit trail is timing — append-only ledgers and timelines exist today; a system-wide record of who changed what arrives with the persistence layer.
Everything the record touches
Specifications with provenance
Seven sources — manufacturer, vendor, commissioning, site verified, field measured, imported. Change a verified value and the verification clears; nobody inherits confidence from an old number.
Fourteen typed relationships
Stored once and read from either side — stand on the panel and it says “controlled by BMS-01”; stand on the controller and it says “controls”.
Analysis coverage, made visible
Which machines have failure analysis and which have none, ordered so the critical ones surface first — because the gap between “analysed” and “not” is the risk nobody wrote down.
Hidden failures, named
A failure nobody would notice is flagged and tracked separately, and coverage counts as managed only when every one of its modes carries a revealing strategy — not just one of them.
Corrective-action effectiveness
Recommended fixes are one record; the thirty-eight times somebody performed them are thirty-eight more. That split is what makes “does this actually work?” answerable.
Retire, don't delete
Relationships end with a date, strategies supersede, library entries retire. “This pump used to feed that tank” is maintenance history, not a mistake to erase.
Your words, everywhere
Rename Asset, Criticality, Health or any other term and it changes what people read — never what a filter matches on.
Site-scoped throughout
Every list, tree, export and figure respects the sites somebody is authorized for — enforced in the query, not by hiding a menu item.
Two languages, all the way down
Every label, hint, empty state and refusal in en-CA and fr-CA — including the sentence explaining why a figure could not be computed.
Records that can't disagree with each other
The discipline behind this module is that a fact lives in exactly one place and every view derives from it. That is what stops an asset list and an asset record telling two different stories about the same machine five years after anybody remembers which was right.
Referenced, never retyped
Location is one reference into the registry — rename a building and it renames everywhere. The manufacturer lives on the catalog model, so it cannot contradict it.
Evidence with every figure
What was counted, what was excluded and why — attached to the number rather than remembered per screen.
Bilingual by birth
Full en-CA / fr-CA parity on every screen, hint, empty state and refusal — plurals handled properly, not by concatenation.
Accessible to everyone
WCAG 2.2 AA as a build rule: real table semantics, states announced in words, and focus that moves to the field an error refers to.
Before you book the demo
What is the difference between your CMMS and your EAM?
Do you handle depreciation and book value?
Can it build our capital plan?
Do you do RCM properly, or is it a checkbox?
Why is there no RPN or risk score?
Can we import our existing asset register?
Do you predict remaining useful life?
How do you handle assets across many sites?
What happens to history when equipment changes?
Is there an audit trail?
How long before it is worth anything?
Where do we start?
Bring the machine you keep arguing about.
The one somebody wants to replace and somebody else wants to rebuild, where the case for each is a spreadsheet and a memory. Bring it and we'll assemble the record in front of you — what it must do, how it fails, what has been done about that, what it has cost, and what the evidence actually supports.