NEPCON Asia 2026: How to Plan an Electronics Manufacturing Line Sourcing Visit

Electronics-production equipment is easy to admire one machine at a time. A placement system can run quickly in a demonstration, an inspection station can produce an impressive image, and a robot can move material smoothly inside a controlled cell. The factory result, however, depends on how those machines handle the same product, takt, changeover, data and quality decisions as one line.

NEPCON Asia 2026 takes place from 27–29 October 2026 at Shenzhen World Exhibition & Convention Center in Bao’an. Its centre of gravity is electronics manufacturing: SMT and PCBA, soldering, dispensing and coating, test and measurement, electronic materials, EMS, semiconductor packaging and production control. S-Factory Expo Shenzhen 2026 runs on the same dates at the same venue and adds industrial robots, machine vision, system integration, industrial software, intelligent warehousing and logistics. The pairing is useful when a buyer brings one production case through both shows instead of collecting unrelated automation ideas.

Define the board, volume and production problem before the show

A useful line brief begins with the product and factory conditions. “We need an SMT line” is too broad: the right process and automation level change with board size, component mix, product life, batch size, quality requirement and local operating environment.

Prepare one representative product family and record:

board dimensions, panel format, layer and surface-finish assumptions;
smallest components, largest or heaviest parts and special packages;
single- or double-sided assembly and through-hole content;
annual volume, batch size, product mix and expected changeovers;
target takt, shift pattern and available floor space;
solder-paste, adhesive, underfill, conformal-coating or cleaning requirements;
inspection, functional-test and traceability expectations;
existing equipment, software and material-handling interfaces;
labour, utilities, environmental control and maintenance constraints;
launch timing and the main yield or capacity problem to solve.

If the project is a retrofit, bring the current process flow and real loss data: queue time, changeover time, first-pass yield, rework, false calls, downtime and the station that limits output. If it is a new line, state which assumptions are still estimates. A supplier should not be rewarded for presenting a precise answer to an input that has never been defined.

Follow the assembly process through NEPCON Asia

NEPCON Asia’s official programme covers the operations that turn a bare board and components into an assembled and tested PCBA. The buyer can therefore follow the actual route instead of organising meetings only by exhibitor brand.

Start with printing and deposition. Compare stencil, paste handling, printer setup, inspection and feedback as a connected operation. Placement should then be assessed against the real component range, feeder demand, board support, nozzle strategy and changeover pattern—not only the machine’s maximum placement rate. For reflow or selective soldering, record the product, alloy, thermal mass, atmosphere, profiling and recipe-control assumptions behind the claimed result.

Continue through the processes the product actually needs:

Process question	Evidence worth discussing at the fair
Dispensing or coating	Material viscosity, deposit or thickness range, keep-out areas, curing, inspection and cleaning
AOI, SPI or X-ray	Defect library, coverage, false-call handling, programming time and feedback to the process
Electrical or functional test	Access strategy, fixture, software, cycle time, limits, data storage and failure diagnosis
Rework and repair	Component and package limits, thermal control, operator guidance and record linkage
Marking and traceability	Unit identity, board genealogy, material lot, recipe, operator, result and rework history

Ask each supplier to place its equipment inside the same line diagram. What enters the station? What information and material does it require? What leaves it? Which upstream variation can it tolerate, and which downstream system uses its result? This reveals hidden interface work that a standalone machine quotation may omit.

Compare line output rather than headline speed

Nominal throughput is only one part of capacity. A line producing several board families can lose more time to feeder preparation, programme change, first-article approval and material replenishment than to the machine’s processing cycle. A fast station can also move the bottleneck to inspection, test or manual handling.

Build a simple capacity model with each vendor. Use the same product, shift length, breaks, changeover frequency, planned maintenance and realistic operating efficiency. Record cycle time by station and identify the constraint. Then test a second scenario: a smaller batch, a denser board or a new package that changes inspection or reflow conditions.

For high-mix production, ask how offline programming, feeder setup, material verification, recipe release and first-article confirmation reduce changeover risk. For high-volume production, focus more on stability, replenishment, predictive maintenance, redundancy and recovery after a stop. Where labour reduction is a goal, calculate which manual task disappears and what new technical work is introduced; automation often shifts work into programming, integration, maintenance and exception handling rather than removing it completely.

Use S-Factory for the movement and control between stations

S-Factory becomes relevant when the unresolved questions sit between machines or beyond the SMT line. Its six official areas include machine vision, industrial IoT and big data, industrial software, robots, system integration, and intelligent warehousing and logistics.

Bring the same line diagram into the concurrent show and mark four types of interface:

Physical flow: boards, trays, reels, components, work-in-progress, finished units and rejected material.
Information flow: order, product revision, material lot, recipe, inspection result, test record and release status.
Decision flow: who or what stops the line, allows a substitute, releases a recipe, diverts a defect or authorises rework.
Exception flow: missing material, damaged carrier, failed scan, machine alarm, network loss, quality hold or urgent order change.

A robot or automated guided vehicle should be discussed against payload, carrier, route, docking accuracy, traffic, battery strategy, safety and recovery. A vision system needs the part presentation, lighting, field of view, resolution, cycle time and acceptable error. Manufacturing software needs more than a dashboard: clarify the source of each data field, system of record, equipment protocol, recipe and user permissions, offline behaviour and ownership of integration work.

System integrators should explain which interfaces are standard, which are customised and who supports them after commissioning. If a machine vendor and software vendor each assume the other will provide a connector, the gap usually emerges late. Record named responsibilities before asking for one combined proposal.

Decide when the concurrent automation show adds value

Not every NEPCON visitor needs a separate S-Factory workstream. A buyer replacing one inspection station on a stable line may gain more from deeper technical meetings and sample testing within NEPCON. The automation extension becomes worthwhile when the project includes several stations, manual transfer, material errors, fragmented traceability, flexible assembly, warehouse-to-line movement or a wider factory digitalisation plan.

The same distinction applies to team composition. A process engineer can compare soldering, placement or inspection detail, while an automation engineer examines handling and controls. IT or manufacturing-systems staff should join when data architecture, permissions, cybersecurity or ERP/MES interfaces affect the decision. Procurement can keep commercial assumptions consistent, but it cannot resolve an undefined engineering interface by asking for a lower package price.

When two workstreams are used, schedule a short reconciliation at the end of each day. Update one line diagram and one open-issue list. A change in carrier, takt, board identification or inspection disposition can invalidate an automation proposal made only hours earlier.

Use demonstrations as controlled engineering discussions

The organiser has announced live demonstration lines for SMT/DIP assembly, semiconductor packaging and test, high-voltage wiring harnesses and product reliability testing. These are useful opportunities to see process sequences and ask specific questions, but the demonstration product and conditions must be recorded.

For a shortlisted machine, request a test plan using the buyer’s own board, component, material or defect samples after the fair. Define the result before the trial: cycle time, repeatability, coverage, false-call rate, changeover, data export, operator intervention and acceptance criteria. For coating or dispensing, include material conditioning and cure. For inspection, include known-good and known-defect samples. For handling, include normal flow and recovery from a blocked or failed station.

Separate the commercial quotation into equipment, options, tooling, fixtures, feeders, software licences, integration, installation, training, acceptance support, spares and recurring costs. Ask which items are required for the demonstrated result. A lower base-machine price is not comparable if the working configuration depends on unpriced options or engineering.

Reserve the third day for system-level closure

A three-day visit can follow the maturity of the decision:

Day one — map the process: walk the relevant NEPCON sections in production order, validate the line brief and identify serious station-level candidates.
Day two — test interfaces: hold deeper technical meetings, observe relevant live lines and take unresolved movement, vision, software or logistics questions into S-Factory.
Day three — close the system: revisit the strongest suppliers with an updated line diagram, compare responsibilities and agree the sample tests, factory meetings and quotation revisions required after Shenzhen.

By the end of the show, each candidate should sit in one of four positions: technically promising and ready for a controlled trial; commercially interesting but missing interface evidence; unsuitable for the defined product and line; or relevant to a future phase rather than the current project. That is more useful than a long list of “potential suppliers” with no next decision.

Leave with one line model and one responsibility map

The final working file should connect process, equipment and automation. It needs the product assumptions, station sequence, cycle-time model, bottleneck, inspection and test strategy, material flow, traceability fields, software interfaces, utility needs and open risks. Beside each item, name the party responsible for evidence, integration, acceptance and support.

NEPCON Asia is strongest when used to understand how electronics are assembled, inspected and controlled. S-Factory adds value where that manufacturing route depends on robots, vision, software and internal logistics. The two shows do not remove the need for sample trials or factory acceptance; they make it possible to design those next steps around one coherent production system instead of eight unrelated machine conversations.

Sources

Event information checked on 30 July 2026.

Xentra can help electronics manufacturers prepare a bilingual line brief, arrange focused equipment and integration meetings, compare quotations on a common basis and coordinate post-show sample trials or factory visits in South China.