KEYENCE PLC Migration in Thailand: RFP and Acceptance Guide

When a Thai factory upgrades a KEYENCE PLC, replacing the CPU and downloading a ladder program is not the finish line. The real shutdown risk depends on whether the source is authoritative, retained values and unit parameters are recoverable, connections to sensors, vision systems, robots and MES can be reproduced, the plant can roll back safely, and local maintenance can own the next change.

This is not a general product overview. It is a practical guide for converting a KV Series upgrade, KV STUDIO migration and EtherNet/IP integration into a procurable RFP, FAT/SAT protocol, shutdown plan, acceptance criteria and source handover. A series name alone is never treated as proof of compatibility. The exact models, software and function versions, topology and application behavior must be established through official documentation and testing.

Safety note: Going online with a PLC, downloading a project, changing RUN/PROGRAM mode, forcing I/O or modifying network settings can start or stop machinery. Only qualified personnel should work under an approved method statement, energy-isolation and machine-safety controls, verified backups and named authorization. Test counts and decision times in a project are risk-based design values, not universal requirements imposed by KEYENCE, ODVA or a standard.

KEYENCE, KV, KV STUDIO and EtherNet/IP are trademarks or registered trademarks of their respective owners. This independent procurement guide does not imply manufacturer endorsement, partnership or a compatibility guarantee for a particular machine.

Accept reproducible control, not merely a new CPU

The objective is to reproduce the safety, quality, cycle, communication and maintenance conditions that existed before the migration, while making every intentional change explainable. Acceptance therefore covers at least six asset groups:

1. CPU, power supply, I/O, specialty units, networks, HMI and connected-device hardware;
2. Ladder logic, variables, comments, labels, unit settings, communication settings, alarms and dependent libraries;
3. Retained values, recipes, offsets, origins and counters needed to resume production;
4. Signal and communication contracts with sensors, servos, robots, inspection systems and MES;
5. Start, stop, setup, automatic, abnormal and recovery behavior; and
6. The as-built source, licences, credentials process, drawings, spares, training and change history needed to maintain the machine.

A matching ladder comparison does not prove that a recipe, device value or external-device setting was preserved. Producing one good part does not prove emergency-related interfaces, communication loss, restart, material shortage or reject handling. The RFP must say which states will be reproduced and which evidence will establish acceptance.

KEYENCE provides official software, manuals, FAQs, connection guides and training for the KV Series. As checked on 2 September 2026, the KEYENCE Thailand installer page listed a Global edition of KV STUDIO Version 12. The published release will change over time, and installing the newest software does not establish compatibility for every installed CPU, project or third-party device. Recheck the official page and the manual for the exact model immediately before engineering work, and open a copy of the project on a validation workstation before touching production.

Step 1 — Freeze the exact models, versions and connections

Before requesting quotations, expand “KV Series” into a verified bill of hardware, software and interfaces. Do not rely solely on an old drawing or purchase order. Reconcile nameplates, the KV STUDIO project, online information, unit configuration and physical network within an approved inspection window.

Tag every photo with the machine, panel and date. Similar panels in one factory are easily confused. Network drawings should include not only IP addresses but also the owner, VLAN, switch port, traffic direction and consequence of an outage.

A filename does not identify the master source

Names such as final, latest2 or backup_new are not configuration control. For every candidate project, record the save time, custodian, source workstation, target CPU, KV STUDIO release, file hash and comparison result. Where safe and approved, compare a snapshot read from the production PLC with the controlled source candidate.

Do not automatically declare the PLC or the engineering PC to be correct. The PLC may contain an emergency edit that was never documented; the PC may contain an unapproved change prepared for the next shutdown. Review each difference for production use, approval and test status, then consolidate one signed baseline.

IEC 61131-3 does not guarantee project portability

The IEC lists IEC 61131-3:2025 Edition 4 as the current edition; the 2013 edition is withdrawn. Product statements concerning standards compliance can be useful when evaluating structure and language, but they do not promise that project files, vendor instructions, device addresses, tasks, motion or communication settings will convert automatically between tools or CPUs. Confirm migration using the official information for the exact product and an application-level difference test.

Step 2 — Build a backup set that can actually restore production

Backup quality is measured by successful restoration from another controlled environment, not by a “save completed” message. A KEYENCE FAQ describes project save/load approaches for backing up and restoring variable data on applicable systems. The exact supported CPU, software release, included objects and restoration conditions must still be checked in the relevant official manual.

The pre-migration recovery set should include:

• the controlled KV STUDIO project and an immutable recovery copy;
• CPU, unit, motion, network and communication parameter exports;
• variables, labels, comments, function blocks, scripts and dependent libraries;
• retained values, recipes, offsets, calibration and counters needed for restart;
• approved backups of the HMI, robot, servo, vision, sensor and gateway;
• I/O, wiring and network drawings, sequence charts, alarm and interlock matrices;
• installed software, version, driver, installer source and licence entitlement;
• online comparison, hashes, extraction time, engineer and approver; and
• restoration and rollback procedures, required tools and prerequisites.

Store the controlled copy and a recovery copy in separate protected locations. Do not leave passwords in plaintext beside the project. Do not depend on a former employee’s email, a personal drive or a licence held only by the integrator.

Rehearse restoration before the shutdown

Use a validation workstation or spare CPU where practical. Install the specified software and driver, open the delivered project using only the handover instruction, confirm the target model, save the check/compile result and run the available simulation or bench test. Verify that comments, variables and all unit parameters are visible.

This exposes missing libraries, unavailable installers, machine-bound licences, unknown passwords, misunderstood device-value handling, orphaned HMI or robot projects, and I/O slots that differ from drawings. Any one of these first discovered during a production shutdown consumes the rollback window.

Step 3 — Write a KEYENCE PLC upgrade RFP that bidders can price

“Replace the old PLC with the latest model” is not a comparable scope. One bidder may price hardware only, another may include conversion, panel modification and site commissioning, while neither includes source handover. Require deliverables and verification evidence.

Separate engineering, hardware, panel work, software change, FAT, transport, installation, SAT, out-of-hours work, training, documentation, spares and post-warranty support. Make assumptions and exclusions explicit. Decide who bears the impact of inaccurate legacy drawings, extra I/O, third-party firmware, network changes or unavailable production material.

Define Thai PLC technical-support boundaries

The manufacturer, machine builder, system integrator, plant maintenance, IT, production, safety and quality teams have different responsibilities. Product support does not automatically include ownership of a machine sequence, another vendor’s robot, MES integration, panel wiring or process qualification.

Your RACI should cover model selection, program conversion, I/O checking, IP allocation, machine-safety validation, product trial, backup custody, licence ownership, holiday calls and interpretation. Confirm Thai/English/Japanese coverage, business and after-hours windows, remote versus on-site terms, factory entry requirements, managed laptops and imported-spares arrangements. The KEYENCE Thailand support page is an appropriate official route to product resources and local contact details, but project-specific coverage and response must be contracted in writing.

Step 4 — Control every KV STUDIO migration difference

A KEYENCE FAQ describes use of Model Change for a KV-7000 project moving to a KV-8000 and identifies a minimum KV STUDIO version for that FAQ case. This is a useful starting point, not an unconditional guarantee for the complete machine. After model change, review unsupported or semantically changed items, unit configuration, allocation, scan/task behavior, retained data, HMI references and every external dependency.

Maintain a migration register:

A clean compile is only an entry gate. A bit may represent a maintained completion, a one-scan pulse or a latched request waiting for acknowledgment. Those require different tests. Include timer/counter limits, first scan, power-cycle retention, manual/automatic transitions, interrupted cycles, re-entry and reset preconditions.

Avoid combining an uncontrolled software refactor with a hardware migration. First establish functional equivalence; then treat improvements as separately approved changes. Review blocks, naming, multilingual comments, unexplained constants, state transitions, timeouts, retries, alarms and interlocks. For a Thai maintenance team, important signals and recovery conditions should use a consistent English/Thai vocabulary rather than Japanese-only comments.

Step 5 — Turn PLC EtherNet/IP integration into a communication contract

ODVA defines EtherNet/IP as an industrial network using standard Ethernet and TCP/IP technologies with CIP. Having “EtherNet/IP” on both datasheets is important, but it does not prove that the required application data and recovery behavior match. The scanner/originator and adapter/target roles, EDS, assembly instances, I/O sizes, data layout, byte/word order, RPI, timeout, multicast/unicast choice, IP method and firmware must be agreed.

Create one controlled contract for every connection:

ODVA guidance encourages testing in application-relevant systems and scenarios, with peer interoperability, cable disconnects and high network load among exception conditions. Device conformance matters, but it does not replace the factory’s application test. Use a representative device or accurate simulator in FAT and the actual managed network in SAT.

Test stale data and validity, not just normal values

After a link is lost, the last normal value may remain visible in memory. Carry communication status and data-valid information with the process value, plus a source timestamp or update counter where appropriate. The machine sequence must not treat stale data as a new command.

Test normal cyclic exchange, peer-not-ready, cable removal, switch reboot, connection timeout, PLC-only reboot, peer-only reboot, revision mismatch and reconnection. For each case establish the safe state, work-in-process disposition, alarm clarity for local operators, and protection against duplicated commands after reset.

Step 6 — Use FAT to prevent defects travelling to the factory

FAT is an acceptance test against the RFP, not a sales demonstration. Every test case needs a purpose, precondition, input, procedure, expected result, actual result, evidence, tester, witness, date and exact hardware/software revision. Link screenshots, trends, logs, videos and measurements to the tested source revision.

At minimum, FAT should demonstrate:

• the project opens and checks on a clean controlled workstation;
• all I/O points can be exercised on a simulator or panel bench;
• setup, manual, automatic and maintenance modes and transitions;
• start, stop, pause, cycle stop and safety-related interfaces;
• stuck sensors, incomplete actuators, timeout and invalid data;
• alarms, interlocks, reset preconditions, history and operator messages;
• EtherNet/IP normal, abnormal and reconnect behavior;
• startup state, retained data and recipes after a power cycle;
• restoration from the delivered recovery set; and
• explainable old/new timing for critical sequences.

Machine-safety validation must be performed by qualified parties against the applicable risk assessment, regulations and standards. Do not infer a safety-function guarantee from standard PLC logic. Classify open FAT defects by severity, owner, due date, workaround and eligibility for SAT. Safety, product-quality, equipment-damage and nonrecoverable issues should remain shipment or cutover gates.

Step 7 — Combine SAT, shutdown and rollback in one cutover book

SAT must be designed before site arrival. Production, maintenance, quality, IT/OT, the integrator and machine builder need one runbook that identifies who performs each task, when it happens and which evidence authorises the next gate.

Go/no-go before stopping production

Confirm the signed baseline, FAT result, unresolved defects, replacement and spare parts, tools, licences, recovery backup, old hardware, network approval, safety permits, production material, quality authority and escalation tree. A missing critical prerequisite should trigger an explicit risk decision, not an automatic start because a date was booked.

Execute the cutover with decision gates

A typical sequence captures the last good product and work-in-process, isolates energy, takes the final backup, labels wiring, removes the old configuration, installs and checks the new configuration, restores parameters, proves I/O, runs dry cycles, verifies communications and interlocks, trials production and obtains quality approval. Adapt the sequence to the site risk and formal procedures.

The timed plan must include rollback decision points, not only optimistic task durations. Continuing because a fault “looks almost fixed” can consume the time needed to reinstall the old configuration. Set decision times from business impact, technical restoration duration and safety—not from a generic template.

SAT covers point-to-point I/O, every relevant mode, representative recipes, normal and boundary cycles, alarms, communication loss/recovery, power cycle, product quality, cycle behavior, operator use and maintenance diagnostics. Keep “the machine moves” separate from “the product is accepted,” and require signatures from production and quality.

Measurable criteria may include complete records for every approved I/O point, expected safe response for all critical interlock cases, first-piece acceptance for the agreed products/recipes, no duplicated actuation or undetected stale value during communication recovery, matching source and PLC revisions, and zero unresolved top-severity defects. Sample size and cycle tolerance must follow the process risk and customer requirements.

Step 8 — Make source handover and local competence completion gates

Avoid discovering after payment that the integrator owns the source, the licence exists only on a loaned PC or a departed employee holds the password. Define background intellectual property separately from project deliverables, and obtain the files and usage rights required for legitimate backup, recovery, maintenance and modification.

The handover package should contain:

• as-built KV STUDIO source, readable exports, revision history and hashes;
• CPU, unit and network settings, I/O lists, drawings and BOM;
• approved HMI, robot, servo, vision and gateway backups;
• software/OS/driver inventory, installer source and licence responsibility;
• roles and accounts, removal of temporary access and remote-service transfer;
• backup, restore, compare, download/upload, replacement and diagnostic methods;
• alarms, interlocks, sequences, communication maps and known limitations;
• FAT/SAT evidence, closed issues, warranty date and support contacts; and
• spare recommendations and the planned backup/restore drill.

Training should end in demonstration, not attendance. Local maintainers must locate the master, open it offline, confirm the revision, collect a backup, restore on a bench, diagnose an alarm, read communication status and escalate with useful evidence. Establish one glossary for RUN, PROGRAM, interlock, bypass, forced value, retained, rollback and as-built across Thai, English and Japanese procedures.

Decide whether to deliver internally or engage a specialist

Judge the decision by downtime risk and accountable capability, not the number of ladder rungs. Consider external technical support when the plant cannot confidently establish the exact baseline; restore the project, runtime data and third-party systems; reconcile electrical drawings and field I/O; create process-aware tests; diagnose EtherNet/IP failure behavior; integrate safety, quality and production acceptance; make a timed rollback decision; or hand over in the local working language.

Assess bidders through relevant CPU, unit, network and industry examples; their survey method; sample FAT/SAT deliverables; bench capability; Thai site response; source-handover policy; safety competence and subcontractor control. “KEYENCE experience” without this evidence is not a complete qualification.

For the broader procurement approach, see our PLC replacement and retrofit guide for Thai factories. To define outsourced logic and deliverables, use the PLC programming outsourcing RFP and acceptance guide. Where the upgrade includes a panel modification, our control-panel manufacturing and acceptance guide helps align electrical and software responsibility.

Conclusion — procure recoverability, evidence and ownership

A successful KEYENCE PLC migration procures more than a CPU. It establishes an exact baseline, a tested recovery set, explainable KV STUDIO differences, a connection contract for every EtherNet/IP peer, FAT/SAT evidence, a viable rollback and local ownership of source, licences and skills.

Start the RFP with the state to reproduce, the evidence that proves acceptance and the conditions that trigger rollback. Use official material to verify models, releases and documented functions; use configuration records and application tests to decide compatibility for the actual machine.

TOMAS TECH can help Thai factories from the assessment stage with RFP definition, backup inventory, I/O and communication mapping, FAT/SAT, cutover and handover requirements. Even if the replacement model and shutdown date are not yet fixed, share the machine context through our English contact page.

FAQ

Can an existing KV Series project be used unchanged after an upgrade?

It depends on the exact old and new CPU, units and KV STUDIO release. A KEYENCE FAQ documents Model Change for a specific KV-7000-to-KV-8000 case and software condition, but this is not an unconditional guarantee for instructions, I/O allocation, retained data, communications, HMI and third-party devices. Create a migration register and verify the application in FAT/SAT.

What belongs in a KV STUDIO migration backup?

Include the project, CPU/unit/network settings, retained values and recipes, dependent libraries, connected-device projects, drawings, installers, drivers, licence information, controlled credentials and restoration instructions. Accept it only after a separate workstation or spare unit can reproduce the required revision.

Will two EtherNet/IP products always interoperate?

No blanket guarantee should be assumed. Match the roles, EDS, assemblies, sizes, data layout, RPI, timeout, firmware and failure behavior. Verify the official documentation for both endpoints and test loss, reboot, timeout and reconnection as well as normal data.

What should we verify in a Thai PLC support provider?

Look for relevant model, unit and network experience; panel and I/O capability; sample FAT/SAT evidence; Thai and English coverage; on-site and after-hours terms; source/licence handover; safe working competence and rollback experience. Put the boundary among the manufacturer, builder, integrator and plant in a RACI.

Do we need both FAT and SAT?

They provide different evidence. FAT removes software, panel, sequence and interface defects before site work. SAT proves the actual wiring, network, machine, product, operator and quality conditions. Design both according to the equipment risk and link every result to a revision.

What is the minimum acceptable source handover?

Require the as-built project and readable export, revision history and hashes, unit/network parameters, connected-device backups, software and licence inventory, drawings, communication mapping, alarm/interlock definition, restore method and FAT/SAT evidence. The plant maintainer should demonstrate opening, checking and restoring the delivered revision.