Solutions

Preventive maintenance

Planned maintenance is the work that stops being urgent the moment anything else is. It slips quietly, it slips first, and the first evidence is usually the failure it was meant to prevent.

  • Biomedical engineering
  • Clinical engineering
  • Quality and compliance
Looking for the other side?Maintenance planningThe scheduling machinery rather than the programme: templates and versions, recurrence, generation ahead of the due date, and what slips.

What is preventive maintenance in healthcare?

Definition

Preventive maintenance (PM)

Preventive maintenance is scheduled work performed on equipment that is still working, on a defined interval or usage trigger, to reduce the chance of unplanned failure and to keep the device within its intended performance and safety envelope.

The distinction that matters operationally is who sets the clock. Corrective work is scheduled by the failure; preventive work is scheduled by you, which is precisely why it is the work that gets postponed. Nothing external protests when it does.

In healthcare it also carries an evidence burden: the work must be provable, by someone competent, against the right procedure, with the result recorded.

Rydya does not decide your intervals

What interval a device needs, what a test must measure, and what limit constitutes a pass are decisions for your qualified staff, your manufacturers' instructions for use, and the standards you work to. Rydya never invents them. It is where you configure those rules, and it is what makes them happen, chases them when they do not, and evidences them when they do.

Why preventive maintenance slips

Not because teams do not care. Because planned work has no advocate: it competes daily against work that is on fire, and losing produces no signal.

A failed ventilator generates phone calls; one due for planned attention generates nothing. Given a finite team and an infinite queue, the silent work loses every time, invisibly.

A spreadsheet schedule is a record of an intention. It cannot escalate, and it cannot distinguish work skipped deliberately from work forgotten: both look like an empty cell.

And planned work is often organised by task rather than by asset, so asking which assets have overdue planned work requires reconciling a schedule against a register against job sheets. It happens before inspections and never in between.

How Rydya makes planned work happen

You configure the rule once against a template; Rydya generates the work, ranks it, chases it, gates it and evidences it.

  1. 1

    Define the procedure as a template

    A template holds the steps, checks, readings and required competencies, written once and reused across every asset it applies to. Versioning records which revision a task ran against.

  2. 2

    Attach a plan to the assets

    A plan binds a template to assets and to the recurrence you have decided. Rydya generates the tasks from that plan. It does not suggest the schedule, and it does not quietly change it.

  3. 3

    Let the work find the technician

    Generated tasks land in a ranked worklist rather than a calendar nobody opens. Due and overdue are evaluated continuously in the background, so overdue is a state the system can act on.

  4. 4

    Do the work where the equipment is

    Technicians work from a phone: the task, its checklist, the readings, the photos. Approved field actions work offline with visible sync state; safety-critical gates stay online by design.

  5. 5

    Prove it without assembling anything

    Completion writes measurements, results, evidence and the transition itself to an append-only, hash-chained audit trail in the same transaction. Compliance is then a report you run, not a fortnight you lose.

How a rule becomes evidenced work

You supply the template and the interval. Everything to the right of that is what the system does with them.

A flow with two inputs on the left, both yours: a versioned template of steps and checks, and your chosen interval, labelled "you decide, not Rydya". Both feed a plan bound to assets, which generates tasks. Tasks lead to two states: due or overdue, in amber, evaluated continuously and escalating; and done and evidenced, in green, with readings, an audit trail and a certificate.Plan + intervalYour ruleTasks generatedOn scheduleDone or overdueNothing forgotten
Note where the boundary sits: the interval is an input on the left, supplied by you. Rydya never generates it, only acts on it.

What you get that a spreadsheet cannot give you

A spreadsheet can hold a schedule. It cannot enforce one, chase one, or prove one.

Overdue is a state, not an opinion

Due and overdue are evaluated continuously, with notifications raised against them. Nobody has to notice, which is what turns a schedule from intentions into pressure.

Escalation without a supervisor in the loop

Unacknowledged or slipping work escalates on rules you configure, so slippage becomes visible to someone who can act while it is still a scheduling problem.

Compliance by asset, not by task

Because plans bind to assets, "which equipment has overdue planned work" is a view rather than a reconciliation project. That is the question inspectors ask and the one spreadsheets answer worst.

The procedure and the record are the same object

The task carries the template version it ran against, the readings captured, who did it and when. There is no separate job sheet to lose.

Deferrals are decisions, not gaps

When planned work has to move, that is a recorded decision with an owner rather than an empty cell that looks identical to work nobody did.

Preventive, corrective and calibration work compared

Different jobs with different triggers and different evidence, and conflating them produces numbers nobody trusts.

How the three kinds of equipment work differ in trigger, timing and evidence
AspectPreventiveCorrectiveCalibration and safety testing
Triggered byA schedule or usage rule you configureA fault report from anyoneA schedule, or a return-to-service requirement
Equipment stateWorking, and stays in serviceFailed or degraded, often out of serviceWorking, but must be proven within limits
Runs toA versioned template of steps and checksDiagnosis, then whatever the fault requiresA procedure with measured readings and pass or fail limits
Main evidenceCompletion, readings, who and whenThe fault, the work done, the return-to-service clearanceMeasurements against limits, certificates, restrictions
Fails quietly?Yes, and that is its defining riskNo, the failure announces itselfYes, until an audit or an incident

How the three kinds of equipment work differ in trigger, timing and evidence

Who owns preventive maintenance

Ownership usually sits with biomedical or clinical engineering, but the schedule only survives if it is visible to the people whose plans it disrupts.

The team that performs planned work rarely controls access to the equipment. A scanner cannot be taken down without radiology agreeing, and a ward will always have a reason why today is inconvenient. Preventive maintenance is a negotiation, and negotiations need a shared record.

That is why Rydya puts planned work in the same system as faults, downtime and cost. When a department can see that the window it refused turned into an unplanned failure, and what that failure cost, the next negotiation is different.

Best practices worth adopting

Operating practices, not clinical rules. What the interval should be remains your decision.

Write the procedure once, reuse it everywhere

A template applied across every matching asset improves the procedure in one place. Per-asset job sheets leave best practice in one technician's version and nowhere else.

Bind plans to assets, not to people

A schedule attached to a person leaves with them. A plan attached to assets survives staff turnover.

Record the deferral, always

A recorded deferral with an owner and a reason is a defensible decision. An empty cell is indistinguishable from negligence, and it will be read as negligence.

Let overdue escalate to someone who can clear the blockage

Escalating overdue work to the technician who is already overloaded achieves nothing. Escalate it to whoever can free the equipment or the time.

Measure completion against the window, not the month

Work completed inside its intended window is compliance. Work completed in the same calendar month is a statistic. Only one means anything in a review.

Questions

How often should we service our medical equipment?

Rydya will not answer that, and you should be wary of any vendor that does. Intervals depend on the device, the manufacturer's instructions for use, your standards, your usage and your risk assessment, and they are decisions for your qualified staff. Rydya is where you configure whatever you decide, and what then enforces, chases and evidences it.

What happens when planned maintenance is overdue?

Overdue is a state the system evaluates continuously rather than something a person has to notice. Overdue work surfaces in ranked worklists, raises notifications to the responsible roles and the assignee, and escalates on the rules you configure.

Can technicians complete planned maintenance offline?

Yes. Approved field actions work offline on mobile with visible sync state, and queued work syncs when the connection returns. Quarantine and return to service stay online only by design, because those gates must be authoritative rather than eventually consistent.

How is preventive maintenance different from calibration?

Preventive maintenance keeps a working device working, against a procedure you define. Calibration proves a device is within limits, measuring against pass or fail criteria, with certificates and restrictions. Rydya runs both, in the same record as the faults and the downtime.

Can we import our existing maintenance schedule?

Yes. Assets and reference data import from a spreadsheet on one screen, with every row validated before anything is written, so bad rows are excluded rather than blocking the rest. Templates and plans are then configured against the imported assets.

See it on your equipment

Live in an afternoon, useful the same week. A person replies, usually within one working day.

Contact us