Key takeaways
It has to produce a per-asset history that stands up as evidence. Concretely, for the failed machine and for a period the adjuster chooses, usually the preceding two to five years: every scheduled job with its due date and its actual completion date, so deferrals are visible rather than erased; who performed each job, by name; what instruction the job followed, meaning a linked procedure or checklist rather than a free-text note; every reactive job and its recorded cause; parts consumed, so a claim about a non-genuine component can be answered; and attached certificates for statutory inspections against that asset.
All of it has to leave the system as a document. Fabrico holds documents and full history per machine, tracks planned versus completed work with dates and the assigned person, ties part consumption to work orders, and exports analytics and job data to Excel. An audit log records who changed what, which is what turns a record into evidence rather than an assertion.
Machinery breakdown cover, sometimes sold as equipment breakdown or engineering insurance, sits apart from property and business interruption cover because the insured event is the failure of the machine itself. That makes the condition of the machine directly relevant to the claim in a way that it is not when a roof blows off.
Insurers manage that exposure in two ways. Before cover, through a survey and through policy conditions that require the equipment to be operated and maintained in accordance with the manufacturer's recommendations and applicable statutory requirements. After a loss, through an adjuster who reconstructs how the machine was looked after in the run up to the failure.
The reconstruction is where plants get hurt. The adjuster is not usually looking for negligence in the dramatic sense. They are looking for a pattern: a preventive task that was scheduled monthly and last completed fourteen months ago, a lubrication route signed off by a name that left the company, a vibration reading trending upward with no corresponding work order. Any of those turns a straightforward claim into a slow one.
The important structural point is that the burden of producing the evidence sits with you. A policy condition can require you to keep records and to produce them. If they do not exist, no one else will make them for you, and reconstructing them after a loss looks exactly like what it is.
None of this is legal advice and policy wordings differ considerably. Read yours, and if the maintenance condition is drafted as a condition precedent to liability, take that to your broker and understand what it means before you need it.
1. The planned regime and its execution. The schedule itself, plus completion against it. The metric that matters is preventive maintenance compliance for the specific asset, not for the plant. A plant running at 92 percent compliance can still have one critical machine at 40 percent, and it is always that machine that fails.
2. Reactive work and causes. Every breakdown, with a cause recorded from a controlled list rather than free text. Repeated small failures on the same component before a large one is a story the adjuster will construct anyway. It is better told by your data than by theirs.
3. Statutory inspection certificates. These live under separate regimes depending on where you operate and what the equipment is. Pressure systems, lifting equipment, electrical installations and, in Germany, equipment falling under the operational safety rules with inspections by an approved body, all generate certificates with expiry dates. Attach them to the asset, and put the expiry in a field that can generate a task, so renewals are driven by the system.
4. Parts and their provenance. Which part went in, and where it came from. Policies and warranties often distinguish between genuine and equivalent parts, and consumption tied to the work order is the only clean way to answer that a year later. Fabrico ties consumption to work orders and holds a supplier listing against parts.
5. Condition and readings. Where you take vibration, thermography, oil analysis or simple gauge readings, they belong against the asset with a date. Readings kept in a separate spreadsheet are the ones that go missing, and they are also the ones that most strongly demonstrate a competent regime.
Question 7 is the one an adjuster's technical advisor would ask. A system where a completed job can be silently edited produces records that are worth less than a system where edits are logged. Ask to see the log, not just to be told it exists.
Fabrico is a combined CMMS and OEE platform. The record-keeping side covers a machine registry with files and full history per asset, an annual preventive maintenance plan with recurring templates and conditional tasks, an auto timer and an emergency and reactive flow that captures reaction time, approval workflows, and an audit log. Inventory covers catalogue, minimum and maximum levels, deliveries, consumption tied to work orders and stock-takes, with suppliers listed against parts.
Analytics cover downtime, MTTR and MTBF, task distribution, top problem machines and per-employee productivity, with Excel export and a custom report builder available through the Fabrico Copilot add-on. On the machine side, availability, performance and quality are calculated from PLC data, which produces a downtime record that nobody typed in. That distinction matters in a claim, because machine-generated data is harder to dispute than a technician's recollection.
Setup on the CMMS side is quoted as 3 days of Fabrico-side work covering configuration, users, roles and bulk import, with bulk-import support included. Machine connection for OEE is a separate exercise paced by the equipment itself. Support response is contractually under 2 hours. The platform is ISO 27001, ISO 9001 and ISO/IEC 20000-1 certified and hosted in an AWS EU region with daily backups and a 4 hour recovery time and recovery point objective, which is also the answer to the question of what happens to your evidence if the system has a bad day.
One honest note on scope: Fabrico is not an insurance or claims system, and no CMMS is. What it does is hold the operational record in a form that can be exported and defended. The claim itself is your broker's territory.
Take a gearbox on a critical line that fails catastrophically. The adjuster requests the maintenance history for that asset for 24 months. Here is the difference between two plants with the same maintenance quality and different systems.
Plant A, spreadsheets and a shared drive. The PM schedule is a workbook with a tab per line. Completion is a colour fill. Nobody can say who filled which cell. Oil analysis reports are PDFs in a folder named by date, not by asset. The last three oil changes are evidenced by an invoice from the lubricant supplier, which proves oil was bought, not that it went into this gearbox. Assembling a response takes two people a week, and it contains gaps that are visible precisely because the assembly was manual.
Plant B, one system. The asset page shows 24 scheduled jobs, 23 completed with dates and names, one deferred with a recorded reason and a completion 9 days late. Four oil analyses are attached with dates. Two reactive jobs are recorded, both with causes, one of which triggered a follow-up job that was completed. Parts consumed include the supplier. Export takes minutes and reads as a continuous story.
The maintenance was equally good in both plants. Only one of them can prove it. That asymmetry, between doing the work and being able to evidence the work, is the entire commercial argument for the system, and it is worth far more at claim time than any efficiency saving quoted in a sales deck.
The evidence discipline here overlaps with formal asset management, covered in our guide to CMMS software for ISO 55001 asset management, and with the way criticality should drive which assets get the deepest regime, in asset criticality analysis. If your plant carries classified areas, the certificate tracking problem is sharper still, and is covered in maintenance software for ATEX hazardous areas. For the metrics an adjuster is implicitly reading, see the OEE for manufacturing guide.
It depends entirely on the policy wording and the jurisdiction, which is why this is a question for your broker rather than for a software article. What is consistently true is that a maintenance condition appears in most machinery breakdown wordings, that the insured is expected to be able to evidence compliance with it, and that claims where the records are strong settle faster than claims where they are reconstructed. Treat the records as the thing you control, and the wording as the thing to check now rather than later.
Longer than you think, and longer than the asset. Adjusters commonly look back two to five years, statutory regimes have their own retention expectations, and a component that failed may have been installed a decade ago. Since digital records cost almost nothing to retain, the sensible default is to keep them for the life of the asset plus a margin, and to confirm with your insurer and your legal function whether anything in your sector requires more.
They help considerably, and they are underused. A photograph attached to a completed job, showing the condition found and the work done, is far more persuasive than a line of text, and it takes a technician seconds on a phone. The requirement is that the photo is attached to the specific job on the specific asset, not sitting in a phone gallery or a group chat.
Yes. The external body issues the certificate, but you are the one who must produce it, prove it was in date at the time of the loss, and show that any defects it raised were actioned. That last part is the one that gets missed: an inspection report with observations creates work, and the evidence that the work happened is your record, not theirs.
It can, in a specific way. OEE data captured from the control system gives an independent, machine-generated record of how the asset was running before the failure, including downtime events and their duration. That is useful both for demonstrating that the machine was not being run abnormally and for supporting a business interruption element, where actual production rates before the loss are the basis of the calculation. It is a secondary benefit rather than a reason to buy, but it is real.
If you want to see a full single-asset history exported end to end, book a demo and ask for the asset history export specifically.
Last updated: 7 August 2026.
Login
Blog