top of page
Newspapers

AUTHENTISE NEWS

Find all of Authentise's press releases, dev blogs and additive manufacturing thought pieces right here.

How to Identify and Fix Additive Manufacturing Bottlenecks

Nov 12
7 min read

How to Identify and Fix Additive Manufacturing Bottlenecks

To identify additive manufacturing bottlenecks, track when work becomes ready, when each operation starts and finishes, and why it waits. Compare the workload arriving at each stage with its available capacity, then investigate recurring queues. The constraint may be the printer, but it could also be preparation, finishing, inspection or an approval that repeatedly holds up delivery.


If your machines look busy while completed orders remain stubbornly slow, that distinction matters. Buying another printer may help - or simply deliver more parts to an already overloaded finishing bench.


A useful diagnosis follows the work from released requirements to accepted parts. Here is how to build that picture and choose an improvement that affects delivery.


Production planner reviewing trays of metal parts waiting for dimensional inspection in an additive manufacturing workshop.


Start with the output you need

Before examining individual machines, define the problem you are trying to solve.

Are you missing delivery dates, struggling to accept more orders or seeing lead times vary unpredictably? Those symptoms may have different causes.


Choose a representative part family or production route and establish:

  • The required quantity and delivery frequency.

  • The operations each part needs.

  • The equipment, people and suppliers involved.

  • The point at which the part counts as accepted and ready for delivery.


Measure output at that final point. A completed build is an intermediate result if the parts still need machining, inspection and release.


Keep unlike routes separate initially. A polymer tool requiring a visual check and a metal component requiring several finishing operations place different demands on production.


For the broader transition to repeatable production, our guide to scaling additive manufacturing to production covers part suitability, process controls, staffing and economics. Bottleneck analysis helps establish which part of that route currently limits performance.


Separate processing time from waiting time

“Cycle time” can mean different things across departments. Agree the start and finish points before comparing figures.


The following distinctions make the investigation easier:

Measure

What it tells you

Processing time

Time required to perform the defined operation, including unattended processing and required cooling or curing where included in that operation

Queue time

How long ready work waits before an operation starts

Other waiting or hold time

Time awaiting prerequisites, decisions, transport or release outside the defined processing operation

End-to-end lead time

Total elapsed time between your agreed starting point and accepted completion

A long printing operation may be entirely predictable. A shorter inspection operation with a growing queue may have a larger effect on delivery reliability.


Record the reason for waiting. “Awaiting inspection” could mean the inspector is occupied, the measurement equipment is unavailable or the required drawing has not been released. Each needs a different response.


Required cooling or curing is part of the manufacturing route. Record it consistently so it is not mistaken for avoidable waiting or counted twice.


Build a timeline for representative jobs

For each stage, record when the work became ready, when it started and when it finished. Add the resource used and the reason for any delay.


Useful stages might include engineering release, build preparation, material release, printing, post-processing, inspection and final documentation. Use the actual route rather than forcing every part through the same template.


In AM, build-level and part-level records need to stay connected. Several orders may share a build and then follow different downstream routes. One build-completion timestamp cannot describe everything that happens afterwards.


Collect information over enough representative work to see a pattern. Include ordinary jobs, priority changes, unsuccessful builds and rework where they occur.


Talk to the people doing the work alongside reviewing timestamps. An operator may explain that a machine marked “available” cannot start because the next build plate is still being prepared.


A simple shared record can support the first investigation. The important thing is consistent definitions and enough detail to explain what happened.


Look for a recurring constraint, not just a large queue

A queue is a useful clue. It is not a verdict.

Parts may accumulate before heat treatment because work is deliberately batched. Another queue may appear after an unusual rush order and disappear the following day. Neither observation, on its own, establishes a persistent capacity problem.


Look for repeated combinations:

  • Ready work accumulates before the same stage.

  • That stage struggles to keep up with incoming workload.

  • Downstream operations wait for its output.

  • Its delays repeatedly affect final delivery.


Examine the age of waiting work as well as the quantity. A small number of jobs may have been bypassed repeatedly while newer work moves through.


Check whether the queue contains genuinely ready work. Parts awaiting an engineering decision should be distinguished from parts ready for an available machining slot.


Where approvals create the delay, investigate ownership and workload. An approval taking ten minutes can still hold up production for days if nobody knows who must perform it.


Compare workload with usable capacity

Capacity depends on the work being processed and the resources actually available.

For each stage, compare the workload due in a period with the capacity available during that period. Include setup, changeovers, maintenance, staffing and other commitments.

Use realistic production timings. Nameplate machine capability or an uninterrupted trial run may not represent the working week.


For a simple illustration, imagine one part family with the following weekly demand and sustainable stage capacities:

Item

Parts per week

Required accepted output

40

Printing capacity

60

Finishing capacity

45

Inspection capacity

30

Assuming every part follows this route and passes each stage, inspection limits sustainable output to 30 parts per week. Printing capacity is already sufficient for the required 40. Adding another printer would not resolve the inspection shortfall.


These are illustrative figures. In your own assessment, account for the job mix, yield and the time different parts consume.


For mixed production, resource-hours may be more informative than part counts. Ten straightforward components can create less inspection work than one complex component.

Also include shared resources. A technician, furnace or measurement system may support several production routes, each of which appears manageable when considered alone.


Interpret printer utilisation in context

Utilisation helps explain how equipment is used, but it does not identify the bottleneck by itself.


Define the measure first. Is it printing time divided by calendar time, scheduled operating time or staffed time? Does it include setup or failed builds? Different definitions produce different answers.


A printer with spare time may be waiting for prepared work, released material or an operator. The question is why that time is unused and whether the missing output is needed.


Conversely, a printer can run continuously and still contribute to poor delivery performance if it produces the wrong priorities or feeds a downstream queue faster than it can clear.

Look at what the equipment produces as well as how long it runs. Accepted output and delivery performance give utilisation a purpose.


Trying to maximise every machine’s activity can create unnecessary work-in-progress. Where a downstream stage is constrained, releasing more work upstream may increase waiting without increasing completed orders.


Match the fix to the cause

Once the pattern is clear, choose an intervention that addresses it.


When work is missing information

If jobs wait for drawings, material status or decisions, improve the handoff. Define what must be available before work is released, who supplies it and who resolves an exception.

Make updates accessible to the people who need them. If the next shift spends its first hour reconstructing the previous shift’s progress, our guide to additive manufacturing shift handovers explains how to carry forward job status, holds and outstanding actions.


When an operation lacks capacity

Check whether setup losses, competing work or avoidable interruptions are consuming the available time. Preparing tools and instructions before the operation starts may help protect a constrained resource.

If demand still exceeds demonstrated capacity, assess additional staffing, equipment, shifts or an appropriate external supplier. The right answer may involve investment.


When work arrives in disruptive batches

Review release timing and batch size against downstream requirements. A large build may use the printer efficiently while producing a wave of work that finishing cannot absorb.

Consider the effect on the complete route before changing the build plan. Splitting work may introduce extra setup or processing effort, so compare both effects.


When rework consumes the capacity

Trace the source of repeated failures or corrections. Rework occupies resources while delaying accepted output.

Resolving the underlying process issue may release capacity. Simply scheduling more tightly around the rework leaves the cause in place.


Test the change against completed output

Choose a defined intervention and agree how you will judge it.

If inspection capacity is the constraint, a trial might reserve a regular inspection slot, prepare the required records earlier or change how work is prioritised. Record what changed so the result can be understood.


Compare similar work before and after the intervention. Useful measures include:

  • Accepted output per period.

  • End-to-end lead time.

  • Queue time and queue age at the affected stage.

  • Delivery performance.

  • Rework and first-pass acceptance.

  • Additional staff effort or operating cost.


A shorter queue is encouraging, but check why it shortened. It may reflect better flow—or fewer jobs reaching the operation because of a new upstream problem.

Watch the next stage too. Increasing finishing output can expose an inspection constraint that was previously less visible.


The improvement needs to help the production route deliver what customers require, rather than merely improve one department’s figures.


Where workflow software helps the diagnosis

Connected production records can make it easier to see where work is, which operation is next and what is holding it up.


Authentise FlowsAM connects scheduling, machines, materials and production records. That provides a useful basis for investigating delays across jobs and production stages.

The value comes from connecting status to meaning. A machine signal may establish that equipment stopped; an operator update or quality record may explain whether the next operation can begin.


Define statuses that distinguish “ready”, “in progress”, “on hold” and “complete”. Capture the reason for important holds and assign responsibility for resolving them.

Software can support the investigation and the resulting coordination. If the evidence shows that a furnace or inspection resource lacks capacity, the improvement still needs to address that resource.

Let the evidence guide the next investment

The printer may be your bottleneck. It may also have spare capacity while work waits elsewhere.


Choose one production route, establish when jobs become ready and follow them through to accepted completion. Compare recurring queues with workload, capacity and the reasons for delay.


That gives the team a practical basis for deciding whether to improve a handoff, protect scarce capacity, change release timing or invest in additional resources.

Start with the stage that repeatedly holds up the required output, make a targeted change and measure what happens to delivery.

bottom of page
window.addEventListener("message", function (event) { // Inspect event.data in console to see what the bot emits if (!event.data) return; // Log all messages to see what your specific bot emits upon submission console.log("Chatbot message received:", event.data); // Check if the event matches a lead/submission event from the bot // (Replace 'form_submit' with the exact event name your chatbot emits) if ( event.data.type === "form_submit" || event.data.action === "lead_submitted" || JSON.stringify(event.data).includes("lead") ) { window.dataLayer = window.dataLayer || []; window.dataLayer.push({ event: "chat_interaction", element_id: "chatbot", form_id: "chatbot", action: "form_submission" }); } });