An A3 report makes the chain from problem to facts, cause, countermeasure, and follow-up visible on one page. In maintenance it helps several functions inspect the same decision logic and convert learning into controlled orders, standards, and reviews.
The worked example follows delayed preventive inspection orders for belt conveyors in a shipping area. The plant expects 42 orders per week, yet some cross the planned week, find equipment unavailable, or return because access and basic material were not ready. All figures and names are instructional rather than customer results.
The completed page at a glance
| A3 block | Conveyor case content |
|---|---|
| Background | Inspections protect critical components before the weekly shipping peak; delays shorten the time available to treat findings. |
| Current condition | Across eight weeks, 336 orders were expected. 214 finished within the week, 74 finished late, and 48 were rescheduled without execution. |
| Target | At least 85% weekly completion for six consecutive weeks without extra overtime, schedule breaks, or evidence-free closure. |
| Supported cause | Readiness review happens after scheduling; access, release, and kits are discovered near execution. |
| Countermeasures | A 72-hour readiness checkpoint, grouping by access window, a verified standard kit, and a separate critical inspection operation. |
| Implementation | A four-week pilot on 12 conveyors with two crews, named owners, and daily reviews. |
| Follow-up | Weekly review of completion, break causes, overtime, evidence quality, and discovered anomalies. |
This summary can fit a visual page because each sentence and chart serves a decision. A living A3 starts incomplete, absorbs evidence, loses disproven hypotheses, and records changes. Extracts, extensive lists, and photographs stay in appendices. The page keeps what a reviewer needs to follow and challenge the reasoning.
Background: earn the capacity spent on the problem
Shipping depends on the conveyors during concentrated operating windows. A late inspection can reveal a roller, scraper, or tracking anomaly too close to the peak. The team then compresses planning, access, and material preparation. Delay therefore affects risk treatment as well as an administrative order metric.
The background establishes relevance, customer, process, and horizon without prescribing a solution. Statements such as people lack discipline or we need an app are removed because they conceal the workflow and bias observation before the current condition is understood.
The maintenance manager sponsors the A3 and the planning coordinator owns the process. Operations, inspection, stores, safety, and execution supervisors participate where their routines create decisions. The A3 owner guides the reasoning but cannot assume their authority.
Current condition: show how work actually moves
The map starts with preventive order generation, then scope review, capacity check, scheduling, access confirmation, kit preparation, dispatch, inspection, finding registration, and closure. The team observes one full week and reconciles the flow with eight weeks of records.
Three queues become visible. Orders wait for access confirmation, kits are prepared only after daily dispatch, and field findings return to planning without enough classification. Wrench time is shorter than the elapsed time between decisions. The map displays those waits instead of calling the whole condition low productivity.
Of 336 expected jobs, 214 finish within the week, or 63.7% in the instructional dataset. Seventy-four finish later and 48 are rescheduled. Among the 122 outside the week, 51 cite access or release, 32 material or tools, 21 priority changes, and 18 other reasons. Text classifications are reviewed because early free-text entries mix delay causes with job descriptions.
Data quality appears on the page. Twenty-seven reasons were reconstructed through interviews and receive a different marker. Nine lack sufficient evidence. The main chart shows counts and proportions, while a note states the limitation. Difficult cases are not deleted to improve the baseline.
Target condition: define success and its boundaries
The approved target is at least 85% of pilot inspections completed in the planned week for six consecutive weeks. The sentence includes population, calculation, duration, and persistence. It does not promise zero rescheduling because production risk and emergency work will still change priorities.
Three balancing measures accompany completion: overtime, changes after weekly schedule freeze, and orders closed without minimum evidence. A higher rate achieved by extending shifts, reducing inspection scope, or confirming early would fail to create the desired condition.
The 21.3 percentage-point gap describes distance rather than cause. The A3 avoids cascading targets to every crew while the common readiness process remains the main subject.
Causal analysis: trace decisions that arrive too late
The team maps the 51 access-related delays. Weekly scheduling closes on Thursday, but the operating window may be confirmed only the day before work. Stores prepare the kit when the supervisor issues the daily list. An order can look scheduled while two critical decisions remain unresolved.
The insufficient-capacity hypothesis is tested against weeks with similar volume and better completion. Available hours exist, but they are assigned to work later found unready. Low technician productivity also loses support because execution duration for completed inspections remains near plan.
The selected cause is a late readiness checkpoint: access, operational release, kit, tool, and skill have no owner and due time before freeze. The escape lies in the scheduled label, which does not distinguish a truly ready order from one merely placed on a date. This explanation matches observed queues and can be tested.
Not every delay receives the same cause. Genuine priority changes remain on their own path. The A3 selects the largest controllable group within pilot scope instead of pretending one mechanism explains all variation.
Countermeasures: redesign the decision sequence
A readiness checkpoint occurs 72 hours before execution. An order becomes ready only when access, release, kit, tool, and required competency have evidence. Conveyors sharing an operating window are grouped. A replenished standard kit supports common findings. The critical inspection and anomaly record become a separate order operation.
A matrix compares expected impact on cause, effort, risk, reversibility, and learning speed. Larger stock without consumption control is rejected because it does not solve release timing. Additional overtime is excluded as a routine response. The pilot combines changes that move decisions earlier.
| Countermeasure | Cause addressed | Implementation proof | Failure signal |
|---|---|---|---|
| 72-hour readiness check | Late discovery of dependencies | Checklist with owner and timestamp | Release granted with an open dependency |
| Access grouping | Fragmented schedule | Weekly package linked to an operating window | Travel or waiting without available work |
| Standard kit | Preparation after dispatch | Identified kit with replenishment record | Basic shortage discovered in the field |
| Separate critical operation | Acceptance without a result | Operation-level confirmation and finding | Closure with no inspection outcome |
Implementation plan with ownership and learning
The pilot covers 12 conveyors for four weeks. During week zero, planning configures the readiness check and selects orders, stores prepares two kits, operations confirms windows, and supervisors practice on a real order. The A3 owner reviews dependencies daily in week one; routine ownership then moves to the process team.
Every action has a named owner in the working record, date, expected output, and completion rule. Training ends after a participant demonstrates and corrects an order, not after attendance. Configuration ends when a test proves that status and evidence reach the intended role.
The pilot order retains operations, resources, materials, and acceptance criteria. If a dependency fails, the supervisor identifies the unmet readiness condition and chooses ready substitute work or documented rescheduling. Generic delay text would erase the learning.
Follow-up: compare prediction with observed behavior
The weekly session updates completion, schedule-break reasons, and balances. Every countermeasure has a separate implementation check. A readiness check may be active while the outcome has not changed. That difference directs investigation toward adherence, delay, or an incomplete cause.
In week one, 10 of 12 orders finish on time, but two require overtime, so distribution is adjusted. Week two delivers 11, while one is rescheduled for a legitimate operating risk and overtime is zero. Week three completes 10 and loses two kits to failed replenishment, which adds a post-use count. Week four completes 11 with full evidence.
Four weeks do not satisfy the six-week target. The A3 stays open and reports a partial result. Leadership can continue the pilot without relabeling an incomplete exposure as sustained effectiveness.
The reaction rule distinguishes a valid exception from a process failure. Emerging risk allows rescheduling with evidence. A known dependency still open at freeze triggers correction of the readiness rule. Closure without an inspection result follows the local reopening governance.
Link the one-page logic to SAP PM
The A3 references system data without replacing the system of record. Orders provide dates, operations, materials, and status. Confirmations provide progress and actual time. Notifications preserve discovered anomalies. Approved recurring changes enter controlled task lists. An aggregated number on the page must be reproducible from documented rules.
A pilot readiness classification must match the organization's data model. If planning maintains it outside a standard SAP status, the classification remains clearly identified and linked to the order. SAP governance chooses supported implementation and controls change.
Use the maintenance planning and control guide to deepen scheduling, the maintenance backlog guide to understand queues, and the work order template to structure execution.
Design the page for review
The left side holds background, current-state map, facts, and cause. The right side contains target, countermeasures, plan, and follow-up. Arrows link causes to changes. A graph earns space only when it changes a decision. Color has a legend and never hides figures or uncertainty.
The reading sequence is explicit: why it matters, what is happening, what gap matters, what explains behavior, what will be tested, who acts, and how the team will know. A block that cannot answer its question is rewritten or removed.
Queries, samples, minutes, and photographs are versioned as appendices. The page itself carries owner and update date. A screenshot inserted into a slide loses decision history; a controlled A3 shows how conclusions evolved.
Checks against decorative A3 reports
- Current condition accepted: the workflow was observed and figures have sources. Countermeasure design waits if this evidence is weak.
- Cause supported: the explanation matches the pattern and survives alternatives. Otherwise, the plan prioritizes a test.
- Countermeasure linked: each change alters an observable causal condition. Generic activity returns for revision.
- Pilot executable: order, capacity, access, material, safety, and proof are ready before the date.
- Follow-up owned: the process owner understands the measure, frequency, and reaction after facilitation ends.
The precise PM Run role
PM Run is a planning, mobility, and execution layer over SAP PM. Planning can organize capacity, shifts, skills, scheduling, and pilot orders integrated with SAP. Mobility can return findings and evidence associated with completed work. SAP retains technical objects and official history.
The product does not write the A3, choose a cause, or approve a countermeasure. It also does not provide sensors, predictive AI, or a ready-made analytical view. Automated data can support customer-built views, while decision authority remains with planning, engineering, operations, and leadership.
Questions for the follow-up room
Must an A3 stay on physical A3 paper?
The paper size created a useful synthesis constraint, but a controlled digital page can preserve the same visible logic. It must retain version, ownership, and appendices without expanding into an endless report.
How much detail belongs on the page?
Keep data needed to understand or challenge the decision. Hundreds of orders, calculations, and raw evidence stay in appendices. Any short statement that carries a conclusion should identify its supporting source.
When should a task list change?
After a recurring countermeasure has been tested and accepted by the technical authority. The A3 records rationale and plan; controlled change occurs in the proper SAP PM object under company governance.
How does the team keep the metric from becoming the problem?
Continue observing the workflow, sampling individual cases, and using balancing measures. Weekly completion shows an outcome but cannot reveal whether access, kits, or legitimate priorities caused the movement.
Technical sources
- ASQ, A3 Report, professional reference for the problem-solving structure.
- NHS England, A3 problem-solving toolkit, neutral guidance on current state and countermeasures.
- SAP Help Portal, Maintenance Order, official documentation for planned maintenance work.
