Uptime and MTTR KPIs for Installed Smart Locks: A Buyer and Operator Guide
Installed smart locks create a measurement problem after handover: the facility team has working doors, incident channels and service records, but no agreed way to separate a short operational interruption from a repair that required several visits. This uptime and mttr kpis operator guide provides a post-award operating procedure for a facility manager. It defines the clock, denominator, incident states, closure evidence and review routine that connect reported KPIs to the door schedule. Use it to establish a baseline, not to turn an assumption into a promise. The buyer sets the target; the operating team supplies the measurement. Supplier support and product-specific handling remain subject to written confirmation.
Direct answer. The one-line answer — uptime and mttr kpis operator guide is the working record for uptime and mttr kpis operator guide. uptime and mttr kpis operator guide is worthless until uptime-and-mttr-kpis is attached to it as evidence. An uptime and mttr kpis operator guide is the operating-life view. It starts after the invoice. uptime and mttr kpis operator guide is the first interval of uptime and mttr kpis operator guide to measure.
uptime and mttr kpis operator guide means here what uptime and mttr kpis operator guide sets out. It is a working definition of uptime and mttr kpis operator guide for this market and product. It is drawn from the evidence below, and it is not a general marketing phrase.
Set the operating scope before setting a target
Start with the object schedule, not the lock count. For every installed smart lock in scope, record the door or unit identifier, site, model and variant from the label or catalogue row, operating criticality, scheduled operating period and response owner. The object schedule becomes the denominator for door-level availability and the population for MTTR. A room label can change; the identifier must not. For any model-specific field, confirm on the dated product page for the specific SKU before using it as a service assumption. The catalogue entry point for this is the smart door locks.
Define scheduled time before the first incident arrives. It is the interval during which the door is expected to be available according to the buyer’s operating plan; it is not simply the time a device spends connected to an application. A door that is not commissioned, is in a planned replacement or sits outside its scheduled period can be excluded only when the reason and dates are recorded. A door with an active fault remains in scope even before a technician attends.
Target fields belong to the buyer. Record each approved target, its approver, review period and version in a controlled KPI sheet. If no target has been approved, report the observed result with its population, period and exclusions rather than inserting an assumed value.
| Scope field | Required record | Control |
|---|---|---|
| Object identifier | Stable door or unit ID | Prevents duplicate populations |
| Site and cohort | Site, building and door group | Makes comparisons meaningful |
| Model and variant | Label or catalogue value | Supports grouping after verification |
| Scheduled period | Expected operating interval | Defines the denominator |
| Exclusion | Reason, dates and approver | Makes changes auditable |
| Ownership | Primary and backup owner | Assigns every incident |
Define the population, states and clock
Use a simple state model so each team describes the same door consistently. An operable door can be opened and secured through the approved route and the site accepts that condition. An impaired door has a reported or observed condition that prevents normal operation. A restored door has been made operable again and has a recorded recovery time. A pending verification door may appear repaired, but closure waits for the agreed check. Define the approved route for each door rather than assuming every site has the same fallback.
Keep the incident timestamps separate. Administrative closure must not replace physical recovery, and an acknowledgement must not be used as a substitute for the recovery time.
| Timestamp | Definition | KPI use |
|---|---|---|
| Report | Fault received through the agreed channel | Starts response and resolution intervals |
| Acknowledgement | A person or system acknowledges the report | Ends the response interval |
| Attendance | A responder reaches the door | Ends the attendance interval |
| Recovery | Door is operable through the approved route | Ends the resolution interval |
| Closure | Required evidence is accepted | Completes the administrative record |
Use one clock and one stated time zone across reports, mobile entries, service-desk records and spreadsheets. Response time is the interval from report to acknowledgement. On-site attendance time runs from acknowledgement to attendance. Resolution time runs from report to recovery. These are labels for measurement, not promised service levels.
If a timestamp is missing, leave it missing and record the uncertainty. Do not estimate an event time merely to make a report appear complete. An incident can close administratively while remaining open in the KPI population when physical recovery has not been verified.
Use a fault record that can produce both measures
Create one incident record as the source for both KPIs. The minimum record includes the object identifier, model and variant, report time, symptom, measured observation, checks performed, parts or unit changes, recovery time and supporting evidence. Model and variant information comes from the label or catalogue, with product-specific details confirmed on the dated product page for the specific SKU.
Write the symptom before writing the diagnosis. “The door did not respond at the approved route” records an observation. “The motor failed” states a conclusion that needs supporting checks. Separate the physical state, indicator or application message, hand movement, power observation and mismatch between physical and recorded states. If a check was not performed, mark it as not checked rather than leaving an inference in the field.
Preserve the sequence of work. If the first visit restores operation but another visit is needed for the same symptom, retain both visits under the incident. If the later symptom is genuinely new, open a linked incident with its own report and recovery timestamps. When a unit is exchanged, record the old unit identifier, its disposition, the parts fitted and the checks made after the exchange.
| Required field | Minimum content |
|---|---|
| Object | Door or unit identifier |
| Product identity | Model and variant with source |
| Symptom | Technician’s words plus measured observation |
| Diagnostic sequence | Checks run and what each ruled out |
| Credential state | Before and after, where relevant |
| Work performed | Parts fitted, unit swapped and old-unit disposition |
| Evidence | Photo, reading or other supporting record |
| Lifecycle | Report, recovery, closure and assigned owner |
A record without the “what each check ruled out” field is a list of actions rather than diagnostic evidence. The fault resolution log and related door operability records should use this same minimum structure.
Run the incident path from report to verification
Follow the same path for every reported fault:
- Capture the report. Open or update the incident through the agreed channel. Record the report time, source, reporter and object identifier without changing the supplied facts.
- Acknowledge ownership. Name the person or service desk responsible for the next action and record the acknowledgement time.
- Describe the impact. Enter the symptom in the reporter’s words, followed by any measured observation. Do not infer a cause from an alert alone.
- Classify the event. Apply the buyer’s severity and continuity rules, then preserve the original classification and any later change.
- Triage before dispatch. Record what can be checked remotely, what remains unknown and which check is appropriate next.
- Attend and diagnose. Record arrival, checks in order and the evidence that supports each conclusion.
- Restore and verify. Confirm that the door is operable through the approved route, then enter the recovery time and verification result.
- Close with evidence. Attach the supporting record, state the resolution, update the KPI dataset and leave the incident open if recovery cannot be verified.
flowchart TD
A[Fault reported] --> B[Create incident record]
B --> C[Record symptom and door impact]
C --> D[Assign owner and severity]
D --> E[Run diagnostic checks]
E --> F{Door operable}
F -->|No| G[Escalate and keep open]
F -->|Yes| H[Verify and record recovery]
H --> I[Close incident and update data]
G --> E
Classify severity without changing the history
Severity determines operational routing; it does not determine whether an incident enters the KPI population. Define the levels against the buyer’s access and continuity requirements rather than against assumptions about a device type.
| Level | Buyer-defined meaning | Required routing |
|---|---|---|
| Critical | Immediate continuity impact at the designated door | Named owner and site escalation |
| Major | Material interruption to the normal operating route | Planned technician response |
| Minor | Limited impairment with an approved route available | Standard work queue |
| Observation | Issue recorded for investigation without impairment | Review and preventive action |
Freeze the initial severity when the incident is opened. If it changes, retain the original value, reason, approver and timestamp. Never move the report time to make the event appear less serious or more serious. Keep severity, operational impact and recurrence as separate fields. A severe event can resolve quickly, while a lower-severity event can remain open for an extended period.
Severity must not alter the uptime denominator or remove an incident from the MTTR population. It helps the team decide what to do next; the measurement rules decide what is counted.
Diagnose at the door in a repeatable order
Use a fixed sequence so that the first check is not skipped because the symptom resembles a familiar issue. The sequence below is a method, not a guarantee of diagnosis. Record the observation and what it rules out at every step.
- Check the physical state. Establish whether the lock is locked or unlocked and whether the handle is free. This helps separate a drive problem from door alignment, closer force or latch preload.
- Check mechanical movement. If the approved mechanical movement turns freely, the seized-lock-body explanation is less likely. Record the actual movement rather than a general conclusion.
- Read the available indications. Capture the indicator and application message. A low-battery indication, a jam indication and no response are different observations with different routes.
- Check power. Test the cells or pack and any emergency input present on the specific unit. A low-battery alert is a threshold, not a measured power reading.
- Compare physical and recorded states. A mismatch between the door and the controller’s state points toward configuration or synchronisation rather than the drive.
- Check the credential route. Where relevant, observe whether the cylinder or fitted reader responds. Record the credential state before and after any intervention.
- Compare the pattern. If the same symptom appears on two doors of the same model, preserve the evidence and investigate configuration, software or supply-batch causes before repeated replacement.
- Replace only after the checks. Record the unit removed, unit fitted, post-work tests and recovery time. Do not erase the failed unit from the incident history.
The Uptime and MTTR KPIs: Which Branch Do You Take decision guide can support the branch selection, while this record supplies the evidence behind it.
CONFIRM-PER-SUPPLIER: Ask the supplier to confirm which observations, diagnostic outputs and supporting evidence are required before it reviews an installed-unit fault or return request.
Calculate installed smart lock uptime
Use installed smart lock uptime as the label for the operational measure only after the scope and denominator are approved. This is lock availability measurement at door level: for each door and reporting period, calculate operable time divided by scheduled time.
Keep every interval in a calculation ledger. The numerator is scheduled time minus periods when the door is not operable through the approved route. Planned work, decommissioning and out-of-schedule periods belong in an explicit exclusion or category field. Unknown intervals remain unknown; they are not quietly treated as operable or excluded.
Calculate each door first, then aggregate with scheduled time as the weight. Do not take a simple average of door percentages, because a door with a short operating period would receive the same influence as one with continuous operation. A site or building result must also state the included population, reporting period, source records and treatment of missing data.
| Calculation field | Operating rule |
|---|---|
| Scheduled time | Agreed expected operating interval |
| Operable time | Scheduled time excluding verified impaired intervals |
| Impaired interval | Report to recovery unless an approved exclusion applies |
| Excluded interval | Reason, start, end and approval recorded |
| Missing interval | Flagged as unknown and excluded from certainty claims |
| Population | Object schedule version used for the period |
Device events may support the record, but they should not replace physical verification when the two differ. The How to Measure and Record Uptime and MTTR KPIs comparison explains the measurement choices behind this calculation.
Calculate MTTR without hiding the tail
Begin the MTTR calculation by defining which incidents belong in the denominator. Use incidents reported or discovered under the agreed population rule within the stated period that reached verified recovery. For each eligible incident, the resolution interval runs from the report timestamp to the recovery timestamp.
MTTR is the mean of those resolution intervals:
MTTR = sum of eligible resolution intervals ÷ number of eligible incidents
Do not substitute acknowledgement time, attendance time, technician arrival or administrative closure. These measure different stages of the incident. Visits may explain the duration, but the visit count is not MTTR. If a report time is missing, keep the incident in an exception view rather than inventing a duration.
The mean also does not describe the tail, the unresolved population or the reason for the delay. Show open-incident aging alongside it. Where the data permits, add median duration, an upper-duration view, repeat rate and first-time fix as companion measures. Their definitions must remain consistent across periods.
The Uptime and MTTR KPIs Worksheet: What to Write Down and Why provides a practical structure for preserving the report and recovery timestamps used in this calculation.
Build the dashboard and review rhythm
Make the dashboard an operating tool rather than a presentation layer. It should answer whether the population is complete, which doors are impaired, which incidents remain unresolved, which faults are recurring and whether every published result traces back to an incident record.
Include these views:
- installed smart lock uptime by approved population;
- MTTR calculation and its eligible incident count;
- open incidents by state and age;
- repeat rate by door and symptom;
- first-time-fix outcome;
- evidence completeness and missing timestamps;
- target variance using buyer-approved targets only.
Use this dashboard as a service-level playbook with clear ownership. The service desk manages the incident queue and missing data. The facility manager reviews operational consequences and exceptions. The supplier contact receives complete evidence requests but does not control the buyer’s denominator. The owner of each measure must be able to explain how a displayed number was produced.
Review the active queue regularly, conduct an operational review of recurring faults and unresolved incidents, and perform a periodic portfolio review of the KPI definitions and object schedule. The Running Uptime and MTTR KPIs Across Several Sites guide expands the governance issue when several sites report into one programme.
For facility management KPIs, keep the definition, population, period, exclusions and source version visible beside the result. A number without those controls cannot be compared responsibly with another site or period.
Separate service outcomes from asset signals
Each KPI answers a different question. Combining them into one score can hide the operational cause of a result and make the wrong corrective action appear successful.
| Signal | What it answers | What it does not answer |
|---|---|---|
| Uptime | How much scheduled door time was operable | Which doors or causes drove the result |
| MTTR | How long eligible incidents took to recover | Whether the tail is long or the root cause was found |
| First-time fix | Whether the visit restored operation without a repeat visit | Whether the diagnosis was correct |
| Repeat rate | Which failures returned within the stated window | How severe each recurrence was |
| DOA | Whether a unit failed before normal service began | Who was responsible for the failure |
| RMA | Which return-authorisation route was used | What the complete commercial process will be |
Planned preventive work and corrective fault work must also remain distinguishable. A preventive check can support future uptime without repairing a reported fault. A corrective visit belongs in the incident record. A no-fault-found outcome is an investigation result and should not be presented as a successful restoration unless the approved operating route was available and verified.
Use the KPI views to direct investigation rather than to assign blame. A rising MTTR calls for examination of the incident path. A rising repeat rate calls for root-cause analysis. A falling availability figure calls for examination of both door impairment and changes to the scheduled-time population.
Handle exceptions and data-quality gaps
Create an exception log so that missing information remains visible. Each exception should identify the affected incident, missing or disputed field, effect on the KPI, owner and decision. Never silently convert a missing value to zero, an operable state or an excluded interval.
| Exception | Required treatment |
|---|---|
| Missing report time | Record the channel receipt time separately and flag the gap |
| Missing recovery time | Keep the incident open or in an ageing view |
| Duplicate incident | Link the records and count the eligible event once |
| Recurring symptom | Preserve separate events and apply the agreed recurrence rule |
| Planned work | Classify it explicitly with approved reason and dates |
| Door removed | End its scheduled population prospectively and retain history |
| No fault found | Record the checks and outcome without claiming a repair |
| Unknown product identity | Mark it unknown; do not infer from a nearby unit |
A fault discovered during an inspection but never reported needs a discovery timestamp and an agreed treatment in the reporting policy. Do not backdate it to resemble a reported incident. The same rule applies when a service record and the physical door disagree: preserve both observations until the approved verification resolves them.
Review exceptions when definitions or the object schedule change. A corrected report time, newly approved exclusion or de-duplication can change a result. Version the calculation and retain the prior result so the effect of the correction is explainable.
Request evidence from the supplier and assign internal ownership
Supplier evidence can support the operating method, but it becomes a KPI fact only when it is incorporated into the buyer’s records. Ask for process evidence, written capability statements and product-specific information tied to the dated page for the installed SKU.
CONFIRM-PER-SUPPLIER: Confirm in writing which timestamps, fault states, diagnostic outputs, export fields and escalation routes the supplier can provide, and which product-specific details apply to the specific SKU.
| Question for the supplier | Evidence to request | Internal owner |
|---|---|---|
| Which clock is used? | Timestamp source and time-zone convention | Service desk |
| Which events define recovery? | Supported state list and event definitions | Facility manager |
| Can incidents be exported? | Door ID, symptom, checks and status fields | KPI owner |
| How is escalation handled? | Written route and evidence requirements | Contract owner |
| What is needed for review or return? | Accepted diagnostic and identification evidence | Service desk |
| Which product details apply? | Dated product page for the specific SKU | Procurement lead |
Assign the internal owner of each KPI, the incident queue and each exception separately. One person may hold several roles in a small facility team, but the responsibility must still be visible. Procurement can obtain supplier evidence; it should not alter operational timestamps. The facility manager can approve definitions; that approval should not rewrite historical records.
Keep this operating method separate from legal or compliance advice. CONFIRM-WITH-LOCAL-AUTHORITY: Verify which authority, adviser or site policy applies before an access, egress or safety requirement enters the KPI scope.
Hand over the operating pack and release it for use
Treat the uptime and mttr kpis operator guide as a controlled operating procedure with a defined release gate. The object schedule should reconcile with site handover or retrofit records, including the Office Smart Lock Site Survey and Smart Lock Retrofit Site Survey where relevant. These records provide scope inputs; they do not become uptime evidence by themselves.
The operating pack should contain:
- the approved object schedule and population rules;
- the KPI dictionary and calculation formulas;
- the fault resolution log and diagnostic checklist;
- the escalation directory and product-specific evidence request;
- the dashboard definition, data dictionary and exception log;
- named internal owners and the review calendar.
Validate the pack by running a recorded incident from intake through recovery and closure. Check that the report and recovery timestamps survive every handoff, that the denominator remains linked to the object schedule and that excluded periods carry approval. Correct the fields before the routine is released; the validation record is not a performance benchmark.
The practical counterpart to this point is smart lock cad drawing requirements. That is the same question uptime and mttr kpis operator guide has to answer here.
| Release control | Acceptance evidence |
|---|---|
| Scope | Approved object schedule and exclusions |
| Clock | One timestamp convention across systems |
| Record | Completed incident passes field validation |
| Calculation | Result reproduces from source records |
| Ownership | Named owner for every KPI and exception |
| Support | Written supplier evidence request issued |
Release the pack when the facility manager can explain the denominator, reproduce a KPI from its source records and identify the next action for every open incident. The finished operating loop is simple: every reported fault has a clock, every restored door has evidence, and every published KPI traces back to the object schedule.
The requirement is ready for request a quote once it is written down.

