The first question in automatic label applicator selection is not the advertised top speed. It is what product receives the label, on which surface, with what data, and how the applied result will be checked. This guide takes a packaging and dispatch line in a Thailand factory as its setting and works through application methods, conveyor synchronization, printing, inspection, quotations and acceptance tests in the order a buyer can use.
This article begins where our guide to the label printing system ends. That guide deals with identifier assignment, print data and reprint control. Here the scope is the equipment that separates a label from its liner, places it on a real item and checks the outcome. A carton side, a cylindrical container and multiple pallet faces are different mechanical problems. The aim is to help production, quality and purchasing teams give several suppliers the same workpieces and the same measurable RFP, rather than to compare robots in the abstract.
Three boundaries to establish before selecting an automatic label applicator
Issuing, printing, applying and verifying are separate jobs
Four activities are often bundled into the phrase “label automation”: assigning an identifier, printing variable information, applying the label to the product, and verifying product and label together. If the RFP treats them as one job, a software quotation may leave the physical application untouched, or an applicator may attach incorrect data more efficiently. The issuance system governs part numbers, lots, dispatch orders and reprint history. The applicator detects a workpiece, peels a label from its liner and transfers it to a specified position. The verification system assesses presence, position, readability and correspondence with the product, then stops or rejects a failed item. State the owner and interface of each activity in the RFP.
Preprinted labels or on-demand print-and-apply
If design and data repeat, an applicator fed by preprinted rolls is a candidate. If lot, serial number, SSCC or destination changes by case, an on-line print-and-apply machine is a candidate. Videojet’s official labeler overview explicitly distinguishes preprinted-label applicators from systems printing variable data on demand. Variable data does not automatically rule out preprinting. A controlled upstream printing process might prepare rolls for each job. Compare the risk of fitting the wrong roll, changeovers and unused labels along with equipment cost and line behavior.
Present, readable and assigned to the right carton are distinct checks
A visible label may contain an unreadable code. A readable code may belong to another carton. A presence sensor does not verify print content; a camera that decodes a code cannot prove the correct workpiece was labeled unless its workpiece ID is captured and compared with the dispatch job. The GS1 Logistic Label Guideline separates visual appearance, data content and barcode technical parameters in implementation verification. If a shipment uses a GS1 logistic label, agree the applicable customer and GS1 requirements before specifying the reader. Our article on GS1 Digital Link and QR migration covers the identification side. Adding a 2D symbol and replacing an existing code are different decisions.
Describe the workpiece and label with measured ranges
Before selecting a machine, record not only photographs but dimensions and variation. Collect workpiece length, width and height; travel direction; location and flatness of the application face; corners and seams; label width and length; roll outer and core diameter; winding direction; liner and adhesive; and permitted placement error. A “single carton type” may still bulge with its contents, have a tape ridge or arrive with damaged corners. Stretch wrap on pallets can ripple. These details change the gap to the application head, pressure time and sensor position.
Create one row per application: work-in-process tag, retail pack, outer carton and pallet must not be mixed. Include workpiece ID, minimum/typical/maximum size, material and finish, minimum/typical/maximum conveyor speed, pitch variation, target position and tolerance, whether data is printed before application, and how the result is checked. If an extreme value has not been measured, name the person and date for measurement rather than inventing a value. Give bidders a sample set containing the smallest and largest workpieces, bulging or warped cases, seams and damaged edges as well as typical units. A trial based only on ideal samples provides little evidence about the real line.
Specify label stock independently from the machine. Paper and film, coating, adhesive and release liner affect dispensing and adhesion even when labels look alike. If the site has heat, humidity, condensation, movement from refrigerated to ambient conditions, dust or oily surfaces, test the combined material and equipment under those actual conditions. There is no universal temperature or humidity figure that can safely replace the label supplier’s stated range and a plant trial. Also measure the application orientation. A flat carton side, a cylindrical bottle and two pallet faces call for different rollers, belts, tamp pads, air-assisted methods or pallet heads. HERMA’s 211 is explicitly a semi-automatic wrap-around example for cylindrical products; its catalog limits must not be assumed to cover a different workpiece.
Compare the application mechanisms
Contact, tamp, air, corner-wrap and wrap-around
A contact unit dispenses a label into the workpiece path and uses a roller or brush to press it down. Consider it when a flat surface and repeatable conveyor position are available. A tamp pad holds the label, moves toward the workpiece and may suit height variation or intermittent movement, subject to a trial. Air-assisted placement can reach a surface where contact is undesirable, but label flight, air quality, pressure and material compatibility must be tested. Corner-wrap crosses two carton faces; wrap-around follows a cylindrical surface. No mechanism is inherently best for an entire industry. Workpiece geometry, orientation, pitch, speed and placement tolerance decide.
Videojet’s 9560 page shows Direct Apply, Tamp, Front-of-pack and Corner-wrap configurations. These are examples of mechanisms, not a guarantee that a particular model fits a reader’s line. HERMA 500 lists modular options, left- and right-hand construction, label widths, I/O and industrial communication. HERMA eco is presented as an entry-level applicator. Write the requirement first, compare method and supply scope second, and only then check the limits of specific configurations.
Manual, semi-automatic or fully automatic
Daily label volume alone does not decide automation. Frequent changeovers, very different shapes, loose placement tolerance and an operator already stationed there can make a fixture-assisted manual or semi-automatic process sensible. Conversely, a modest volume may warrant automation when shipment errors, customer traceability, staffing on night shifts or awkward manual posture matter. Compare actual labor, rework, scrap, downtime, inspection, training, consumables and changeovers. Mark unmeasured items as unmeasured instead of assigning convenient savings.
In a semi-automatic arrangement an operator normally loads or locates the workpiece while the machine performs some dispensing and application. A fully automatic line coordinates feed, identification, positioning, application, inspection and rejection with conveyance. A quotation that says “automatic applicator” might still require a person to orient every carton. Draw the remaining human steps on the process diagram and include guarding, access and maintenance space.
Rated speed is not line throughput
Catalog figures for labels per minute or conveyor meters per minute apply to the manufacturer’s stated conditions. The buyer needs throughput with its actual carton length, shortest pitch, application orientation, label length, printing time, inspection and rejection. A dispenser may supply a label quickly while a long carton still occupies the application zone. The next carton can then collide with the operation at a short pitch. Empty gaps, stops, restarts, manual jogs and reverse motion matter as much as steady operation.
Give bidders measured peak flow and its duration, minimum pitch, speed variation and stop frequency, not only an average. Choose between encoder tracking and fixed-delay application according to that variation and the position tolerance. HERMA 500 lists master encoder and integration interfaces, illustrating the type of option to ask about. An interface in a brochure does not mean the device already matches the site PLC. Specify the timing and states for trigger, ready, low label, fault, application complete and reset as well as the protocol.
Define the trial instead of asking for one advertised performance number. Run extreme carton sizes at minimum pitch; normal production speed; start and stop; the end of a roll; immediately after a job change; and an unexpected conveyor pause. Agree how many pieces to run and how to record skew, peel-off, double labels, missing labels, unreadable codes and wrong product data. If an operator quietly repairs failed pieces, count that as rework, not good machine output.
Close the loop across printing, application and inspection
For variable printing, identify whether ERP, MES or WMS holds the master value. Having an operator type a part number into a machine template leaves an error entry point even after application becomes automatic. Design a flow that sends a job ID, workpiece ID, print data or approved template ID to the machine and receives the result. Match workpiece and print job one to one; stop safely if the queue shifts. The label printing system guide deals with issuance and reprint governance, but the equipment RFP must still define IDs received and evidence returned.
Consider checking roll or printed data before application and reading the label on the product afterward. A pre-application check will not see a label lost in transit. A post-application scan cannot detect the correct label on the wrong carton if the carton itself was never identified upstream. Reader placement is affected by vibration, reflection, turning cartons, wrinkled labels, transparent film and print density. Test readability with the worst samples, rather than borrowing a brochure’s result.
For GS1 logistic labels, the GS1 guideline identifies SSCC as the required logistic-unit identifier and separates visual, content and technical verification. That does not impose SSCC on every internal factory tag. Work-in-process labels, customer-specific shipping labels and GS1 logistic labels have different rules. Compare the customer requirement and receiving process with the applicable standard before writing a vision-inspection specification.
Define the fate of each bad item. Missing label, duplicate label, unreadable code and data mismatch need different recovery procedures. For each fault decide line stop or automatic rejection, whether the rejected carton may re-enter, whether its label number is voided, who can approve a restart, and how retesting is recorded. Include a full reject lane and a returned carton that re-enters the original queue. False rejection of good cartons also costs time, so retain counts by fault class, not only an overall pass rate.
Site conditions in a Thailand factory
In a retrofit, sensors, pressure rollers, cameras, rejecters, controls, guarding and maintenance access often determine the footprint more than the applicator head itself. A vacant space on a photo may hide a guide rail that covers the label face, a short straight run between existing machines or an operator route. Mark conveyor direction, door swing, emergency stops, roll replacement and maintenance clearance on a dimensioned layout and require a site survey.
List power, compressed air, network and PLC connections in the quotation assumptions. For an air-assisted unit check flow as well as pressure, the safe state during supply loss, filtration and condensate handling. For networked jobs decide what happens when job data is unavailable, where logs live and how clocks are synchronized. Ethernet or Wi-Fi on a product page says little about compatibility with the plant’s security policy or existing PLC protocol.
Operating language is a requirement too. Confirm whether Thai procedures and screens are needed, who approves in Japanese or English, and whether local maintenance staff can interpret fault codes. Use illustrated roll-threading instructions and physical guides to reduce winding-direction errors on every shift. Define spare-part location and the scope of phone and on-site support. “International support” alone does not answer who responds during a night-shift stoppage. Where hygiene or safety rules apply, give the vendor the actual cleaning, chemical and audit requirements. An IP rating on one module does not establish the rating of the printer, camera and whole installation.
What to put in the RFP and how to read quotations
Ask suppliers to make a line function work, not merely quote a machine. Include a workpiece list, label and face, throughput peaks, line drawing, print data, inspection, rejection, signal list, safety, maintenance, languages and acceptance criteria. Attach product and roll samples, photographs of limit cases, and dummy data that reproduces real field lengths and character sets without revealing customer identifiers.
Split the quotation into applicator, feed/conveyor modifications, printer, scanner or camera, rejecter, panel and PLC work, host-system connection, guarding, installation, FAT, SAT, training, spares, warranty and maintenance. Scope and exclusions explain more than the headline price. “Inspection included” may mean a presence sensor only or full code-and-job matching. “Integration included” may mean a trigger wire or a logged job handshake.
Compare lifecycle cost with actual roll and ribbon changes, label material, wear parts, inspections, template changes, changeovers, software support and the need for standby equipment. Do not treat a foreign list price as an installed Thailand price; currency, freight, installation, tax, service and warranty scope differ. A payback calculation should use measured plant downtime, manual time and scrap together with a written quotation, with all assumptions visible.
Ask each bidder to name one rejected application method and why it was rejected. A contact-system proposal is easier to judge when the supplier explains why tamp or air was not chosen—surface flatness, speed and maintenance may be decisive. Give all bidders the same revision of the RFP and samples, and distribute answers to questions to all of them. Different peak assumptions cannot yield comparable quotations.
One evaluation sheet for purchasing, production, quality and maintenance
Purchasing focuses on initial price, production on changeovers and uptime, quality on wrong labels and evidence, and maintenance on recovery. Separate scorecards can produce a cheap machine without the required inspection, or a high-spec machine that needs a specialist for each roll change. Use a joint sheet with columns for mandatory requirement, comparative criterion, test method, owner and evidence. Document who approves any change to the weights or pass conditions.
Watch operators use the proposed machine before purchase. Check screen language, threading direction, liner disposal, jam-clear posture and product-change steps, alongside guarding and interlock requirements. A machine that is easy to run but impossible for the night shift to replenish independently does not meet the operating brief.
Divide responsibility between label supplier and equipment vendor
If label and machine vendors differ, production rolls must match the tested stock in liner, adhesive, coating, winding direction and roll dimensions. Record tolerances, splice treatment and storage conditions in the purchasing specification. Where label rolls have lot IDs, decide how they relate to equipment fault logs. Require the equipment vendor to state precisely which stock, workpieces and speeds were tested. Agree which tests must be repeated if material changes, and whether an alternative stock affects performance warranty.
Narrow the first installation
Introducing many carton shapes, customer formats and lines at once makes failure diagnosis difficult. Start with a stable workpiece, one application face and a manageable customer format; establish the print-to-inspect-to-reject sequence. Expand sizes, faces and host integration in stages. Still document future options for power, communication, mounting and job management before purchasing, so a second line does not force a redesign.
FAT and SAT must prove more than “a label was applied”
FAT is the supplier-side test before shipment; SAT is acceptance after installation. FAT should cover representative and extreme workpieces, different faces, end of roll, job change and stop/restart. SAT adds the real conveyor, network, PLC, work instructions across shifts and rejected-product handling. Use the same core criteria in both, with site-specific additions at SAT, so differences can be diagnosed.
Agree placement tolerance, missing/double/peeling limits, code reading and job matching, fault stop or rejection, queue integrity after restart and how uptime is recorded. A percentage alone is weak when only a handful of samples are run. State sample count, product mix, speed, assessor and record format. Reading a barcode once with a handheld terminal is not equivalent to formal symbol verification. The GS1 guideline separates visual, content and technical checks for this reason.
Record difficult intermittent faults. Define timestamp, workpiece ID, job ID, label ID, inspection result, equipment state and operator reset, together with retention and access. Retain enough to investigate a shipment query without collecting unnecessary personal data. Agree vendor attendance after startup, spare rolls, first-response contact, authority to change thresholds and the conditions for a temporary return to manual labeling. If manual and automatic flows coexist, record which cartons used each path and reconcile the unfinished queue after recovery.
Build acceptance scenarios around likely failures
Normal application alone does not show how the line behaves under stress. Include a roll with a missing label, a roll threaded in the wrong direction, consecutive cartons with the same part number but different destinations, and a conveyor stop exactly at a job-change boundary. A safe simulation under vendor supervision can establish whether the machine detects the state, where it resumes and what happens to unused labels. For each case assess three things separately: whether the error is detected, whether the physical carton is isolated, and whether a record remains. A double label found by a camera still creates a quality problem if rejected and good cartons are mixed in the reject area. Train operators with the same scenarios, linking each screen fault code to a work instruction and an escalation owner for every shift.
Frequently asked questions about automatic label applicator selection
How is an automatic applicator different from a label printing system?
The printing system governs identifier assignment, approved templates and reprint history. The applicator physically attaches labels and coordinates conveyor and inspection. A print-and-apply machine includes printing, but whether it should own the master identifier is a separate design decision. Assign data ownership first.
How should we estimate the cost of an automatic labeling machine?
Compare identical scopes covering conveyor changes, printing, inspection, rejection, controls, integration, safety, installation, FAT/SAT, training, spares and support. Give vendors actual workpieces and peak conditions and compare exclusions. A catalog equipment price is not the local installed total.
When do we choose preprinted labels versus print-and-apply?
Choose according to variable data by item and whether preprinted inventory and changeovers can be controlled. Fixed content points toward an applicator; case-specific SSCC or destination may point toward print-and-apply. Both still need post-application reading and workpiece matching where required.
Will a camera eliminate wrong-label incidents?
Not by itself. Specify viewing angle, lighting, readability, upstream workpiece identification, job comparison and rejection. A correct label on the wrong carton may look perfect to a camera if the carton ID was never captured. Test deliberately wrong samples at FAT and SAT.
What additional checks matter in Thailand?
Measure the site’s heat, humidity and condensation where relevant; check local operating language, shift changes, spare parts and support, PLC/network fit and guarding. Conditions differ by plant, so use site data and sample trials rather than country-wide assumptions.
Summary
Automatic label applicator selection starts with the real workpiece, target face and data ownership. Then compare preprinted and print-and-apply, application mechanism, conveyor tracking and post-application inspection. Rated speed and machine price are meaningful only under those conditions. Put limit samples, peak flow, signal definitions, fault handling and FAT/SAT tests in the RFP to make proposals comparable and avoid disputes after installation.
If you are considering a retrofit in a Thailand factory, contact TOMAS TECH to review workpiece samples, a line drawing and the inspection boundary. An early RFP scoping discussion is welcome.
