Leave Your Message
The Machine Is Ready, but the Factory Is Not
Industry News

The Machine Is Ready, but the Factory Is Not

2026-08-12

A site-readiness gate for power, air, environment, network and commissioning

11 August 2026

Author: Phyllis Zhao  |  Reviewed by: Poemy Engineering Team

Shanghai Poemy Automation Equipment Co., Ltd.

Contents

  • Executive judgement
  • Why this topic fills a real procurement gap
  • The contract needs three capability statements
  • What each decision-maker should demand
  • Power: nameplate voltage is not power readiness
  • Compressed air: pressure without flow and quality is not a specification
  • Temperature, humidity, dust and sanitation are engineering inputs
  • Network and data readiness must precede remote support
  • Civil, layout and material-flow readiness
  • Upstream and downstream interfaces need a failure contract
  • A seven-gate readiness model
  • Diagnose a missed SAT without turning it into an argument
  • FAT and SAT should test different truths
  • Commercial clauses that support engineering reality
  • When the buyer should not purchase yet
  • A 2026–2046 readiness architecture
  • Procurement-ready site-readiness schedule
  • Five inputs Poemy needs for a first readiness screen
  • Frequently asked questions
  • Sources
  • Related Poemy engineering guides
  • Call to action

Executive judgement

A factory should not release a Packaging Line for shipment merely because the machine is finished. It should release shipment only when the receiving site can reproduce the conditions under which the promised result will be accepted.

The most expensive commissioning failures often begin outside the machine. A Flow Wrapper may be mechanically complete while the incoming voltage dips during a fryer restart. A case packer may cycle correctly while plant air pressure collapses when another production area cleans down. A control system may be ready while the buyer has not approved IP addresses, remote-support rules or historian tags. A carefully designed line may arrive before the floor, cable trays, drainage, access route or lifting plan is ready.

These are not minor site inconveniences. They change the test boundary. If factory utilities and interfaces differ from the conditions assumed during design and FAT, neither buyer nor supplier can tell whether a failed SAT result is caused by the equipment, the facility or an undefined responsibility gap.

This paper argues for a contractual Site Readiness Gate between FAT and shipment. The gate does not shift supplier responsibility to the customer. It separates responsibilities so that machine capability, factory capability and integrated-line capability can each be evidenced. That separation gives the CEO a more reliable launch date, the CFO a clearer delay-cost boundary, engineering an auditable interface specification, procurement a usable responsibility matrix, and operations a safer start-up.

1. Why this topic fills a real procurement gap

Poemy's current Industry News library already covers accumulation behind multiple wrappers, FAT and OEE acceptance, multi-SKU design, automation cost and semi-automatic versus automatic case packing. Those subjects help a buyer select and accept equipment.

The missing question is what the receiving factory must prove before that equipment can be installed, energized, connected and tested. Site readiness is often scattered across quotations, electrical drawings, civil notes, meeting minutes and assumptions. The result is a false sense of progress: the machine schedule is visible, while the facility schedule remains implicit.

Site readiness deserves its own commercial gate because it directly affects launch certainty, SAT validity and cash flow.

 The acceptance boundary.png

Figure 1. The acceptance boundary. Purpose: separate machine capability, factory capability and integrated-line capability. Logic source: Poemy engineering procurement model. Suggested position: after the procurement-gap discussion.

2. The contract needs three capability statements

One sentence such as “the line shall run at the agreed speed” is not enough. A robust project defines three related statements.

Capability

Responsible evidence

Typical proof

Machine capability

Supplier within the agreed machine boundary

FAT with approved products, materials, recipes, test duration, fault rules and accepted-output definition

Factory capability

Buyer within the agreed site boundary

Measured utility records, completed civil and electrical works, approved network/security controls, access and staffing readiness

Integrated-line capability

Buyer and supplier jointly

SAT under agreed site conditions with live upstream/downstream interfaces and an agreed stabilization plan

The distinction prevents two bad outcomes. First, the buyer should not accept a weak machine because its factory has imperfections. Second, the supplier should not be held responsible for performance that cannot be tested because the agreed utilities or interfaces were never made available.

The practical instrument is a Responsibility, Interface and Evidence Matrix. Every interface has one owner, one measurable condition, one verification method, one due date and one consequence if it is not ready.

3. What each decision-maker should demand

CEO or owner: protect the market launch

The CEO needs one integrated critical path, not separate machine and construction schedules. The readiness gate should identify which facility items can stop shipment, installation, dry commissioning, wet commissioning or commercial production. A red item must name an owner and a recovery date; “in progress” is not a control state.

CFO: expose delay cost and cash-flow asymmetry

The CFO should ask whether payment milestones reward evidence or calendar passage. A shipment payment released before the site is ready can create storage, demurrage, double handling, warranty ambiguity and idle commissioning labor.

A simple auditable delay model is:

Expected delay cost = probability of delay × (lost recoverable contribution + idle labor + travel/remobilization + storage/demurrage + restart scrap + financing cost).

Each term must state its time window and source. “Lost sales” should include only output that the business could realistically sell and recover. This is a planning model, not a claim that any particular site will incur the modeled cost.

CTO and engineering: control the interfaces

Engineering should freeze voltage, frequency, prospective fault level, earthing arrangement, compressed-air pressure and quality, ambient limits, network topology, safe isolation, upstream/downstream signals, product/material conditions and data ownership before detailed design closes.

Procurement: remove scope-shaped ambiguity

Procurement should not accept phrases such as “utilities by buyer” without quantities, measurement points and tolerances. The contract should say where power quality is measured, where air pressure is measured, which cable and connector ends belong to whom, which party installs network drops, and what happens if SAT cannot begin.

Operations and maintenance: make recovery possible on the night shift

Operators need access, lighting, safe clearing paths, recipe authority, material staging, change parts, cleaning tools and clear escalation. Maintenance needs isolation points, drawings, backups, diagnostic access, lifting points, critical spares and a restoration test. A facility that can run only while the supplier's senior engineer is present is not operationally ready.

4. Power: nameplate voltage is not power readiness

Machine nameplates do not describe the electrical environment. A factory can present the nominal voltage while still exposing equipment to dips, interruptions, imbalance, harmonics, transients, poor earthing or inadequate short-circuit capacity.

IEC 61000-2-4:2024 addresses compatibility levels for conducted disturbances at in-plant points of coupling in industrial locations. It is a useful reference for agreeing the environment, but it does not automatically become the project's contractual limit. The buyer and supplier must select the applicable class, measurement point and acceptance method for the actual plant.

Minimum electrical readiness record

Item

Required project record

Evidence before shipment

Supply

Nominal voltage, frequency, phases, conductor system and permitted tolerance

Single-line diagram plus measured record at the intended point of connection

Capacity

Connected load, demand basis, diversity assumption and spare capacity

Approved load schedule and protective-device coordination review

Quality

Dips, interruptions, imbalance, harmonics and relevant transient history

Logged data over a representative production window, including large-load starts

Protection

Fault level, breaker ratings, discrimination and emergency isolation

Approved study or responsible-engineer sign-off

Earthing

Earthing arrangement, equipotential bonding and measured continuity

Test report with measurement locations

Backup

UPS or controlled-stop need for controls, vision, network and recipe data

Failure-mode test plan; no implied backup unless specified

The measurement window should include the factory's real disturbance pattern: fryer heater switching, compressors, chillers, welders, elevators, large motors and generator transfer. A ten-minute reading taken during an idle shift is not representative evidence.

5. Compressed air: pressure without flow and quality is not a specification

Pneumatic actuators and vacuum systems can behave normally during a standalone test and fail when other plant users draw from the same header. The readiness specification therefore needs four elements: pressure at the machine connection, available flow during the defined production scenario, pressure-dew-point or moisture requirement, and contaminant class where product or package integrity requires it.

ISO 8573-1:2010 classifies compressed-air purity by particles, water and oil. It should be used as a reference only after the project decides which class is necessary at which measurement location. A generic statement such as “clean, dry air” is not auditable.

A defensible air test

  1. Define the maximum simultaneous machine state to be tested.
  2. Define which other factory users remain operating.
  3. Measure static pressure, dynamic pressure at the point of use and the duration of any excursions.
  4. Verify filtration, condensate management and dryer performance under expected ambient conditions.
  5. Record the recovery behaviour after a high-demand event.

The project should also decide who supplies the final isolation valve, local filter/regulator, hose or hard pipe, connection size, pressure monitoring and low-pressure interlock. These small items frequently become commissioning-day blockers.

 Utility acceptance map.png

Figure 2. Utility acceptance map. Purpose: convert vague utility promises into evidence fields. Logic source: IEC 61000-2-4, ISO 8573-1 and Poemy interface practice. Suggested position: after compressed-air readiness.

6. Temperature, humidity, dust and sanitation are engineering inputs

Packaging performance depends on the combined behaviour of product, film, seals, electronics, sensors, adhesives, cartons and operators. Ambient conditions can change that behaviour.

The site specification should state the expected and worst credible temperature and relative-humidity ranges during production, shutdown and cleaning. It should identify airborne flour or seasoning dust, oil aerosol, corrosive cleaning agents, washdown boundaries, pest-control treatments and condensation risks. The supplier should then state the environment for which the selected components, enclosure ratings, cooling and materials are designed.

No universal humidity or temperature number should be copied into a contract without considering the film, adhesive, carton, product and local climate. The correct value is an agreed design envelope supported by component ratings and product/material testing.

Questions that expose hidden environmental risk

  • Will warm noodle cakes enter the wrapper and change seal behaviour over a shift?
  • Can high humidity soften cartons or affect hot-melt adhesion?
  • Can flour or seasoning dust impair photoelectric sensors, cooling fans or cabinet filters?
  • Does cleaning create spray, steam or condensation near electrical enclosures?
  • Is control-cabinet heat rejected into a room that already approaches its design limit?
  • Are rejected packs, film rolls and change parts stored within the same environmental envelope?

Environmental readiness is not satisfied by installing air conditioning after an SAT problem appears. It should be designed, measured and included in the test record.

7. Network and data readiness must precede remote support

A packaging line may need recipe management, code verification, production reporting, remote diagnostics, historian tags or interfaces to MES/ERP. Those functions cannot be left until an IT ticket is raised during commissioning.

NIST SP 800-82 Rev. 3 provides guidance for securing operational technology while addressing performance, reliability and safety. The ISA/IEC 62443 series provides a lifecycle framework for industrial automation and control-system security. The project should use these as governance references, then translate them into site-specific requirements.

Network readiness decisions

Decision

Minimum answer before shipment

Ownership

Who owns PLC/HMI backups, recipes, production data, audit logs and configuration exports?

Addressing

Approved IP ranges, VLAN/zone, switch ports, firewall paths, DNS/NTP and time source

Remote access

Named users, MFA, plant approval, time-bounded session, logging, termination and file-transfer rules

Change control

Who can change code or recipes, how backups are taken, how rollback is proven and who signs the record?

Recovery

Offline restore media, known-good backups, license keys, compatible hardware and a tested restoration path

Integration

Protocol, tag list, direction, update rate, error handling, data quality and acceptance owner

Permanent unmanaged remote access is not a commissioning shortcut. The plant should be able to authorize, observe and terminate support access. Remote diagnosis should never bypass safety controls or substitute for local isolation and verification.

8. Civil, layout and material-flow readiness

The layout is ready only when the equipment can be delivered, installed, operated, cleaned, maintained and later removed or upgraded.

Civil and logistics evidence

  • Final approved layout with equipment, access aisles, guarding, escape routes and maintenance envelopes.
  • Floor loading, levelness, anchor method, drainage and any plinth or trench detail.
  • Door, corridor, elevator and turning dimensions for the largest shipping unit, including packaging and lifting gear.
  • Rigging plan, crane or forklift capacity, lifting points, roof/headroom limits and safe exclusion zone.
  • Cable tray, pipe route, network drop, air take-off and isolation points coordinated with the as-built equipment position.
  • Material routes for noodle cakes, sachets, films, cartons, adhesives, rejected product, waste and finished cases.
  • Safe operator access for film changes, jam clearing, carton replenishment, cleaning and fault recovery.

A layout that fits the machine footprint but blocks a cabinet door or requires an operator to cross a case conveyor to clear a jam is not ready.

9. Upstream and downstream interfaces need a failure contract

The packaging line does not receive an average product stream. It receives gaps, clusters, off-center cakes, broken cakes, misplaced sachets, film joins and restarts. Downstream, it meets inspection, accumulation, case packing, sealing, coding, palletizing and warehouse constraints.

For each interface, define:

Interface field

Example of an auditable definition

Product state

Orientation, spacing, temperature, allowable damage and accepted/rejected status

Normal rate

Stable accepted rate by SKU under defined materials and staffing—not a nameplate-only figure

Disturbance

Maximum expected gap, surge, stop and restart scenario

Signal

Ready, run, starve, block, fault, emergency stop, recipe and reject-handshake semantics

Buffer

Capacity expressed in accepted packs or seconds at the defined net rate, with product-pressure limits

Recovery

Which machine slows or stops, how product identity is preserved and who clears retained product

 This “failure contract” matters more than a clean animation of steady production. It describes how the system behaves when reality departs from the nominal case.

 Readiness stage gates.png

Figure 3. Readiness stage gates. Purpose: link evidence to commercial release points. Logic source: Poemy procurement workflow. Suggested position: before the formal gate model.

10. A seven-gate readiness model

Gate 0 — measurable business need

Confirm the product mix, accepted output, launch date, labor objective, quality controls and expansion intent. If the business cannot define the result, the project is not ready to select architecture.

Gate 1 — interface design basis

Freeze products, films, sachets, cartons, utilities, environment, upstream/downstream boundaries, safety philosophy, network rules and responsibility matrix. Open assumptions must be visible and commercially owned.

Gate 2 — site design freeze

Approve civil, electrical, pneumatic, network, access and rigging designs against the supplier's latest drawings. Design changes after this gate require an impact assessment on cost, schedule, FAT and SAT.

Gate 3 — FAT release

Confirm test materials, quantities, recipes, staffing, test duration, failure scenarios, data capture and deviation process. FAT proves the agreed machine boundary; it does not prove the unbuilt factory.

Gate 4 — shipment readiness

Require buyer evidence that the receiving site, access route, foundations, main utilities and installation resources will be ready by the agreed arrival date. Exceptions need a written storage, preservation and warranty plan.

Gate 5 — energization and dry commissioning

Verify safe isolation, polarity/phase, earthing, air quality and pressure, network segmentation, guards, emergency stops, documentation, backups and no-load sequence before processing saleable product.

Gate 6 — SAT and stabilization

Run the agreed products and materials with real operators and interfaces. Record accepted output, planned/excluded time, rejects, stops and recoveries using the pre-agreed boundary. Close open points through a timed stabilization plan rather than treating one short run as proof of sustained capability.

11. Diagnose a missed SAT without turning it into an argument

When accepted output misses the target, the project needs a fault-isolation method. Blame is not diagnosis.

 SAT fault tree.png

Figure 4. SAT fault tree. Purpose: separate evidence paths when integrated performance misses target. Logic source: Poemy engineering diagnostic model. Suggested position: beside the SAT diagnosis method.

Evidence sequence

  1. Confirm the test boundary, SKU, material lot, staffing and time classification.
  2. Reconcile raw production counts, accepted packs, rejects and case counts.
  3. Check measured power, air, environment and network events against the agreed envelope.
  4. Classify each stop by initiating cause rather than by the machine that finally stopped.
  5. Reproduce the top loss under controlled conditions where safe and practical.
  6. Assign corrective action, owner, due date, retest condition and commercial consequence.

OEE can assist, but only when availability, performance and quality use controlled definitions. A dashboard that changes its ideal rate or excludes inconvenient downtime cannot settle acceptance.

12. FAT and SAT should test different truths

Test

What it should prove

What it should not pretend to prove

FAT

Equipment sequence, safety functions, product handling, recipes, defined fault recovery, controls, inspection and accepted-output capability within the test boundary

The buyer's utilities, upstream process, downstream warehouse or long-term shift performance

Site readiness test

Utilities, environment, civil works, access, network, staffing, materials and interface availability

Supplier machine performance

SAT

Integrated performance with the agreed live site boundary, materials, people and scenarios

Long-term reliability from a short demonstration

Stabilization review

Repeated-shift performance, loss Pareto, training effectiveness, spares, maintenance and recovery

A reason to leave acceptance definitions open-ended

The buyer should not pay twice for the same proof, and the supplier should not use FAT to close site integration obligations. Each test has a distinct question.

13. Commercial clauses that support engineering reality

Readiness certificate

Create a signed certificate listing every shipment-critical item, its evidence reference, responsible person, date and status. A conditional release must name the exception and mitigation.

Remobilization rule

Define what happens if commissioning cannot proceed because an agreed buyer prerequisite is absent—or because supplier work is incomplete. Include notice, standby, travel, rescheduling and cost treatment for both directions. Balanced clauses reduce opportunistic claims.

Preservation and warranty

If equipment will be stored, state packaging, humidity control, periodic inspection, battery/UPS care, lubrication, corrosion prevention, insurance and the event that starts the warranty clock.

Change control

Every post-freeze change should identify the requirement, reason, safety impact, technical effect, cost, schedule, documentation and acceptance impact. Verbal agreement on site is not adequate configuration control.

Retention tied to evidence

Where commercially appropriate, link final retention to clearly defined SAT/stabilization evidence and closure of critical documentation—not to vague “satisfaction.” Local law, financing terms and bargaining position may require a different structure.

14. When the buyer should not purchase yet

Do not issue a purchase order merely to preserve a target date when any of the following is true:

  • Product and packaging formats are still changing without an agreed design envelope.
  • The upstream discharge and downstream case/pallet requirements are unknown.
  • The facility has no approved electrical load, power-quality or air-capacity evidence.
  • The equipment route, floor, drainage or maintenance access is unresolved.
  • IT and OT teams have not agreed addressing, remote access, backups and data ownership.
  • The buyer expects the supplier to solve undefined production instability through extra machine speed.
  • The factory cannot provide test materials, trained operators or a decision-maker during SAT.
  • Commercial milestones would force shipment into uncontrolled storage.

The correct action may be a paid or bounded site survey, interface workshop, material test or brownfield feasibility study. Deferring an equipment order can be the fastest path to a successful project when the alternative is committing to an undefined boundary.

15. A 2026–2046 readiness architecture

A line purchased in 2026 may face new SKUs, films, case counts, coding rules, cybersecurity controls, obsolescence and labor constraints long before 2046. Site readiness therefore should not optimize only for first energization.2026–2046 readiness roadmap.png

Figure 5. 2026–2046 readiness roadmap. Purpose: connect first-installation choices to future upgrades and renewal. Logic source: Poemy lifecycle strategy. Suggested position: before the lifecycle procurement actions.

Design for controlled change

  • Reserve electrical, pneumatic, network and floor capacity only where a defined expansion scenario justifies it.
  • Use documented interfaces and version-controlled backups so modules can be upgraded without reconstructing the entire system boundary.
  • Record component obsolescence and supported software versions in the annual lifecycle review.
  • Revalidate remote-support controls, accounts and recovery media after organizational or technology changes.
  • Maintain a current product/SKU envelope; new materials should trigger focused trials rather than silent assumption.
  • Revisit buffer, inspection and case-packing architecture when upstream rates or channel requirements materially change.

The long-term value proposition is not that one machine will remain unchanged for twenty years. It is that the architecture can be measured, restored and upgraded without losing control of safety, data or accepted output.

16. Procurement-ready site-readiness schedule

The following schedule can be attached to an RFP or used during clarification. Values must be completed for the actual project; blank fields are risks, not flexibility.

Category

Required buyer input

Supplier response

Verification owner

Due gate

Electrical

Supply, quality, fault level, earthing, protection, connection point

Loads, tolerances, protection and isolation requirements

Buyer electrical lead

G2/G4

Air

Pressure, simultaneous flow, ISO 8573-1 class if required, dew point, connection

Demand scenario, connection, local preparation and low-pressure behavior

Buyer utilities lead

G2/G4

Environment

Temperature, humidity, dust, washdown and cleaning chemistry

Permitted envelope, enclosure/cooling and materials assumptions

Quality + engineering

G2

Civil/access

Floor, drains, anchors, maintenance space, door route and lifting

Loads, footprints, centers of gravity, lifting and access envelopes

Project manager

G2/G4

Product/material

Cake, sachet, film, carton, adhesive, coding and variability

Tested envelope, sample quantity and exclusions

Production + supplier

G1/G3/G6

Controls/network

Signals, topology, addressing, data, remote support and recovery

Architecture, protocols, accounts, backups and cybersecurity controls

OT/IT owner

G2/G5

People

Operators, maintenance, EHS, QA, interpreters and decision authority

Training prerequisites, attendance and escalation plan

Plant manager

G4/G6

Acceptance

Test duration, rate, quality, exclusions, failure scenarios and deviations

FAT/SAT procedure and evidence format

Joint steering team

G1/G3/G6

17. Five inputs Poemy needs for a first readiness screen

Poemy can begin a bounded technical discussion without assuming that a new machine is required. The minimum useful inputs are:

  1. Product and packaging formats, dimensions, weights and representative materials.
  2. Target stable accepted outputby SKU, plus the upstream and downstream rate boundary.
  3. Existing or planned equipment list, layout and electrical/pneumatic/network interface drawings.
  4. The top recurring stops, rejects, manual-handling losses or commissioning concerns.
  5. FAT/SAT boundary, site date, factory utility evidence and intended change/expansion scenarios.

The first output may be “no equipment change,” an interface correction, a brownfield retrofit, a single bottleneck replacement or a phased integrated line. That is preferable to quoting an architecture before the boundary is known.

18. Frequently asked questions

Is nominal voltage enough for a packaging-line quotation?

No. The project should also define frequency, phases, conductor and earthing system, capacity, protection, fault level and relevant power-quality conditions at the point of connection.

Which compressed-air value matters most?

No single value is sufficient. Define dynamic pressure and available flow during the agreed simultaneous-use scenario, plus required purity and moisture conditions at the point of use.

Can FAT compensate for an unfinished site?

FAT can prove the agreed equipment boundary under controlled conditions. It cannot prove the buyer's site utilities, civil works, network or live interfaces.

Who owns a failed SAT?

Ownership follows evidence. Use the responsibility matrix, measured utilities, material records, stop classification and agreed test boundary to identify the initiating cause and corrective owner.

Should shipment wait until every site item is complete?

Not necessarily. Shipment-critical items should be defined in advance. A conditional release may be sensible when open items have credible owners, dates, mitigation, preservation and no unacceptable acceptance risk.

Is remote support a substitute for local maintenance?

No. Remote support can speed diagnosis, but the plant still needs safe isolation, trained staff, backups, spare parts and a local recovery procedure.

How long should the stabilization period be?

Use a period that captures representative shifts, products, changeovers and normal disturbances. The contract should define the window and evidence rather than copy a universal duration.

What is the first sign that a site is not ready?

Unowned assumptions. If utility, interface, access, data or staffing questions have no named owner and evidence due date, the risk is already active.

FAQ schema

The publishing team should implement an FAQPage JSON-LD object with the eight questions and answers above. Required structure: context equals https://schema.org; type equals FAQPage; mainEntity contains Question objects; each Question contains one acceptedAnswer of type Answer. The published page text and schema answers must remain identical in meaning and should be revalidated if the article is edited.

Sources

  1. IEC 61000-2-4:2024 — compatibility levels for industrial power distribution systems
  2. ISO 8573-1:2010 — compressed-air contaminants and purity classes
  3. NIST SP 800-82 Rev. 3 — Guide to Operational Technology Security
  4. ISA/IEC 62443 series — industrial automation and control-system security
  5. Poemy Industry News

Editorial note: Standards are cited as authoritative frameworks. Project-specific limits, classes and test methods must be agreed by the buyer, supplier and responsible local professionals; this paper does not reproduce paid standard text or replace local electrical, safety, building or food regulations.

Related Poemy engineering guides

Call to action

Before requesting a machine price, ask Poemy to review the site-readiness boundary. Send the five inputs above through the Poemy contact page or email poemy01@poemypackaging.com. Poemy will separate verified requirements from assumptions and identify whether the next useful step is an interface review, material test, brownfield study or equipment proposal.