The Machine Stopped. The Schedule Kept Moving.
A breakdown becomes a delivery problem when maintenance evidence and production priorities live separately. Recovery needs one shared operational clock.
The short answer
A printing press breakdown becomes controllable when maintenance and production share the same event, affected jobs, quantity state, recovery options and decision deadlines. Stop the schedule from advancing on stale assumptions, but do not wait for a perfect repair estimate before protecting materials, alternative routes and customer promises.
The silent dashboard
The press stops at 11:20. Maintenance opens the machine and begins diagnosis. Operators move the current pallet aside. On the planning screen, the job continues toward its original finish time.
Sales sees green. Finishing expects material at 3:00. Another shift prepares plates for the next job. The physical machine has stopped, but the digital schedule has not received the event.
By the time the delay becomes visible, the factory has lost hours of decision time.
Create one downtime event
Record machine, start time, observed condition, current job, operation, quantity state and safety status. Link maintenance work and production impact to the same event.
The first record need not contain a confirmed cause or repair time. It should establish that the previous schedule assumption is invalid and identify who owns the next update.
Preserve later diagnosis rather than rewriting the first observation. History helps distinguish symptom, cause, action and verification.
Stabilise the job before replanning
Confirm which quantity is good, uncertain, damaged or still in the machine. Protect artwork, plates, substrate and settings. Identify whether output before the stop needs inspection.
If the intervention may change quality conditions, define a restart approval and first-piece check. The traceability pillar explains why quantities before and after a material machine event may need distinguishable history.
This prevents a schedule recovery from creating a later quality investigation.
Estimate in ranges
Early in a breakdown, a precise repair time is often fiction. Give planning a range and next diagnostic milestone.
For example: safe access in 30 minutes, diagnosis by 12:30, likely recovery between 2:00 and 5:00, replacement component decision at 1:00. Update confidence as evidence improves.
Planning can protect alternatives without presenting the worst case as certainty or the best case as a promise.
Calculate the job impact
Identify the interrupted job, remaining run quantity, setup state and downstream critical path. Then identify queued jobs that share the machine, material, tooling or finishing reservations.
The visible downtime may be two hours, but recovery can require cleaning, setup, colour approval and resequencing. Compare the complete schedule effect.
Prioritise by customer commitment, material readiness, route flexibility, setup family and remaining buffer—not by whichever department calls first.
Open alternative routes
Check compatible machines, shifts, plants and approved subcontractors. Confirm capability, format, colour, tooling, customer approval and transport.
An alternative press may save the current job while creating extra setup or changing quality reference. Present trade-offs and cost.
Keep preparations reversible until the repair evidence or decision threshold justifies the move. Starting plates and transport too early can create a second loss.
Connect maintenance priority to production value
Maintenance needs technical severity; production needs customer and capacity consequence. One priority model should consider safety first, then affected critical jobs, alternative capability, downstream risk and repair resources.
This does not mean sales controls maintenance. It means the team sees why one safe repair has greater operational urgency than another.
Planned maintenance should use the same view so protected windows reflect actual demand and capability.
Communicate milestones, not reassurance
Customer service needs a truthful decision schedule: diagnosis underway, alternative route being qualified, revised promise by a defined time.
Avoid repeatedly saying the machine will “probably be ready soon.” Communicate when the next evidence will arrive and which customer option follows.
Internal teams should see the same update. A single recovery log prevents maintenance, planning and sales from circulating different repair estimates.
Verify the restart
Repair completion is not production recovery. Confirm safety, machine condition, test run, quality acceptance and stable output.
Record which component or setting changed, which batch restarted and who released production. Recalculate the schedule using actual recovery evidence.
If a temporary repair limits speed or format, preserve the restriction. Nominal availability should not become another false capacity promise.
Learn across failures and schedules
Review downtime duration, diagnosis delay, parts availability, repeat failure, lost setup, affected jobs, premium cost and delivery impact.
Maintenance metrics such as mean time to repair are useful, but production context explains business consequence. A short stop at the wrong time may be more costly than a longer planned outage.
Use patterns to improve preventive work, spares, training, standard recovery and alternative-route readiness.
A recovery control card
Capture:
1. machine and event time;
2. current job and quantity state;
3. safety and quality containment;
4. diagnosis range and next update;
5. affected jobs and deadlines;
6. alternative routes and approvals;
7. customer communication owner;
8. restart verification.
This gives the team one operational clock while technical work continues.
Frequently asked questions
Integrate planned and unplanned work
Preventive maintenance and breakdown recovery compete for the same technicians, parts and machine access. Show both on one resource horizon.
Protect planned windows using production demand and risk. If work is deferred, record the reason and revised date rather than silently removing it. Repeated deferral increases failure exposure and should become visible to management.
For common breakdowns, prepare standard diagnostic steps, spares, safe access and external support. This reduces time before a credible recovery range can be given.
Preserve the product context of failure
Machine events may correlate with substrate, format, speed, ink system, tooling or job family. Link the active job and settings without assuming they caused the fault.
Compare repeat events across comparable conditions. Maintenance can identify asset patterns while production can improve operating standards. If a component fails after a particular setup sequence, the evidence should be available to both teams.
Do not let the maintenance code become the only story. “Mechanical failure” may close a work order but offers little operational learning.
Manage the queue after restart
Once the press is stable, decide which job runs first. The interrupted job may have the earliest promise, but another job may have perishable material, a protected finishing slot or a more compatible setup.
Present scenarios with displaced consequences. Confirm plates, material, people and downstream capacity before publishing the new order. A repaired machine does not automatically make every queued job ready.
Monitor the first shifts after restart for speed and quality restrictions. Update the schedule from actual evidence rather than the nominal rating.
Review the whole recovery
The post-event review should follow detection, safe stop, diagnosis, parts, repair, verification, resequencing and customer outcome. Identify time spent waiting for information or authority as well as technical work.
Choose a small number of actions with owners: spare availability, sensor, inspection, escalation, training or alternative-route preparation. Verify them during later maintenance or a controlled drill.
The objective is not to eliminate every stop. It is to reduce recurrence and preserve faster, safer decisions when uncertainty returns.
Should the schedule stop automatically when a machine stops?
The affected assumption should become invalid immediately, while planners control how and when downstream jobs are recalculated.
Who estimates repair time?
Maintenance owns technical evidence. Planning translates the range into production scenarios. Neither should invent the other’s answer.
Is preventive maintenance always more important than urgent production?
Safety and asset integrity come first. Beyond that, use risk and evidence to choose the least harmful protected window.
What is the best downtime KPI?
Combine frequency, duration, repeat cause, recovery stability, affected production and delivery impact. One average hides important context.
Put both clocks together
Create a visual recovery board for significant events. Show safety state, diagnosis milestone, repair range, affected jobs, alternative preparations, next customer decision and named owners. Update it from source evidence rather than repeating assumptions in meetings.
Close the board only when the machine is stable, affected quantities are resolved and the revised schedule has been acknowledged downstream. Preserve the timeline for review.
This common view reduces interruptions to technicians because stakeholders can see the next update. It also stops production teams from acting on an earlier optimistic estimate after maintenance has learned something new.
Over time, compare recovery boards across events. Repeated waits for the same part, authority or verification step reveal improvement opportunities that a simple downtime total cannot show.
The maintenance clock measures diagnosis and safe recovery. The production clock measures disappearing options around jobs and customers.
When both clocks share one event, the factory can protect quality, choose alternatives and communicate honestly. The machine may stop. Operational control does not have to stop with it.