Motionwell Automation designs and builds capping and sealing systems in Singapore for pharmaceutical, medical device, food and consumer product lines. Most enquiries reach us as a request for a bottle capping machine in Singapore, and most of them turn out to be a request for a capping station that has to sit inside an existing line rather than beside it. The delivered reference is the GMP filling and sealing platform on projects P23005 and P25026: rotary capping heads driven by Mitsubishi HG-KR servo motors with torque feedback, handling both ROPP aluminium caps and screw-on plastic closures through quick-change capping chuck assemblies, with torque programmable from 0.5 to 5.0 Nm and the torque curve logged for every container. Cap presence is confirmed by a Keyence IL-600 laser displacement sensor measuring cap height, and containers that fail are pushed to a quarantine bin rather than passed on. The machines are designed, assembled and tested at our Woodlands Link facility, and the company has delivered more than 150 special purpose machines since 2014 under ISO 9001:2015 and bizSAFE Level 3.
Where we stand, said plainly before you read further: capping has always shipped here as a station inside a machine, either the closing section of a filling and sealing line, or a dedicated vial capping and uncapping cell (P23005) built around a collaborative robot with a custom end-of-arm tool and pneumatic torque control. We do not sell a catalogue capper off a shelf, and where a proven standard machine covers your closure at your rate, we will tell you that rather than quote around it.
This page covers how closures are actually formed, why torque control is the centre of the machine, how caps get fed and presented, what fails on a capping station and how each failure gets caught. Capping sits immediately downstream of dosing, so it is worth reading alongside the filling machines we build and the inline vision inspection that judges the result. The delivered build is documented in the GMP filling and sealing case study. If you already have container drawings and a closure sample, skip ahead and talk to an engineer.
Which Capping and Sealing Method Fits Your Container?
Closures are classified by how the seal is formed, not by what the cap looks like. The container finish and the closure supplier decide this before any machine question comes up, and getting it wrong is the one mistake that cannot be tuned out later.
| Method | How the seal is formed | Suits | What it needs from you |
|---|---|---|---|
| Screw capping | A chuck grips a preformed cap and rotates it onto the thread to a set torque | Bottles, jars, vials, most liquids | A torque target from the closure supplier, and a container held square against the reaction torque |
| Snap or press-on | A plunger presses the closure over the finish bead until it clicks home | Dairy cups, flip-tops, some medical containers | Force and depth control, because there is no torque to measure |
| ROPP roll-on pilfer-proof | A plain aluminium shell is threaded onto the finish in place by forming rollers, and a second roller rolls the pilfer band under the neck ring | Spirits, syrups, pharmaceutical liquids in glass | Consistent shell thickness and a neck finish within tolerance; the machine forms the thread, so the glass sets the result |
| Crimp seal | An aluminium seal is crimped over a stopper and the vial neck | Injectable vials, sealed after stoppering | A stopper already seated correctly; crimping a lifted stopper seals the fault in |
| Induction seal | An induction coil heats a foil laminate liner inside an applied cap, bonding it to the container land | Bottles needing a tamper-evident barrier under the cap | A compatible container material, a liner matched to it, and correct application torque so the foil is pressed flat during heating |
| Heat seal | A heated tool bonds a foil or film lid to a tray or cup flange | Trays, cups, blister formats | A clean, flat flange; product on the seal area is the usual defect source |
| Ultrasonic seal | Ultrasonic energy welds the film to the container without a hot tool | Films that scorch, filled products sensitive to heat | Joint design done at container stage, not at machine stage |
Motionwell’s delivered work covers ROPP and screw capping with servo torque control on the pharmaceutical platform, and heat seal, crimp seal and ultrasonic seal integration on the tray-format platform that runs 20 to 60 units per minute. Two honest exclusions. We have not built an induction sealing head of our own design, and there is little reason to: the sealing heads are a settled commodity, and the result is decided by the cap, the liner and the application torque. Where a line needs induction sealing we integrate a proven unit, the same way we integrate can seamers and stretch wrappers. And we do not specify your closure or your liner. That belongs to the closure supplier, whose data sheet gives the application torque window the machine is then built to hold.
Why Is Torque Control the Centre of a Capping Machine?
Because torque is the only variable that decides whether the container leaks, and it is invisible after the fact. A cap that is under-torqued looks identical to a cap that is correct. It leaks in transit, three weeks later, in someone else’s warehouse. A cap that is over-torqued also looks correct, and it distorts the liner, strips the thread, or shears the tamper band so the customer receives a pack that reads as already opened.
The same argument runs on our assembly machines with a different fastener. On the SCARA sensor panel line (P23045), electric screwdrivers work to a target torque set per fastener in the recipe, from 0.1 to 2.0 Nm, measured by a rotary torque transducer in the spindle. The PLC records final torque and total rotation angle for every fastener against the panel serial number, and anything outside the window is rejected immediately, because an under-tightened fastener loosens in service and an over-tightened one strips the thread or cracks the substrate. A closure fails in exactly those two directions. The instrumentation is the same idea applied to a cap.
That is why the capping heads on P23005 and P25026 are servo-driven rather than clutch-driven. A friction clutch or a magnetic clutch releases at a set torque and tells you nothing else, and it drifts as it wears. A servo head with torque feedback gives a programmable range, in our case 0.5 to 5.0 Nm, changed by recipe instead of by adjusting a mechanical stop, plus a continuous reading of what the closure did on its way home.
One boundary worth stating. Application torque, which the machine controls, is not removal torque, which is what your customer experiences and what your QC lab measures. The relationship between them depends on the liner, the thread finish and how long the pack sits, so removal torque has to be measured on filled product, not assumed from the machine setting. We build to your application torque window and prove repeatability at factory acceptance testing. The window itself is your closure supplier’s number, validated by your own removal-torque testing.
What Does a Torque Curve Tell You That a Torque Number Cannot?
A final torque figure is one sample from a process that lasted a few hundred milliseconds. The curve is the whole process, and the shape carries the diagnosis.
A correct application has a recognisable signature: near-zero torque while the cap runs down the thread, then a sharp rise as the liner compresses, then the cutoff. Change any part of the mechanism and the shape changes before the final number does.
- Torque that rises immediately, with almost no run-down, usually means the cap started cross-threaded and jammed on the first thread.
- Torque that never rises to target within the expected angle means the thread stripped, the cap is not engaging, or the container is spinning in the holder.
- A rise that reaches target early, in fewer degrees than normal, points at a liner that is thicker than the last batch, or a closure supplier change nobody told production about.
- The same target reached with a shallow, drawn-out rise usually means the liner is missing.
This is why the machine logs the curve per container rather than a pass or fail flag. Combining torque with rotation angle turns a single value into a window, and the window catches faults that any single number lets through. Servo torque signals are free process signatures generally: our 12-station rotary syringe assembly machine (P22068, 15-second cycle) uses force-monitored insertion the same way to reject out-of-band assemblies.
Where the line is regulated, that log stops being process data and becomes a record. A servo torque curve kept as evidence that a cap was applied correctly falls under 21 CFR Part 11, which changes where it is stored, who can alter it and what audit trail follows it. How that split gets decided during validation is on our electronic batch record page.
How Do Caps Get Fed, Sorted and Presented?
Feeding is where capping lines lose their rated speed, and it is consistently under-scoped in quotations. The head can only apply a cap that arrives the right way up, at the right moment, in the right place.
A typical feed chain has four stages. A bulk hopper holds a shift’s worth of caps at floor level. An elevator lifts them into a sorter. The sorter, usually a vibratory bowl with a tuned track and escapement, or a centrifugal sorter for high rates, rejects caps in the wrong orientation back into the bowl and passes correctly oriented ones onto a chute. The chute delivers to a pick-off point where the chuck or a starwheel takes one cap per container.
We build these with vibratory bowl feeders and escapement mechanisms as standard practice on rotary assembly machines, and bowl feeder modification and upgrade with starwheel indexing is regular work here for high-mix production. The engineering reality is worth stating: on our SCARA assembly lines a bowl feeder tuned to an awkward part can approach the price of the arm it feeds, and closures are not exempt from that arithmetic. A feeder that presents a wrong-way part every fiftieth cycle also costs far more than it looks like it should, because every one of those is either a jam or a reject.
Closure geometry decides how hard this gets. Plain screw caps and ROPP shells are close to symmetrical and sort well. Flip-tops with a hinge, caps with an off-centre spout, child-resistant closures with an inner shell, and trigger sprays with a dip tube are all progressively worse, and a trigger spray is not a bowl-feeding problem at all. Its dip tube tangles, so it usually needs a dedicated sorter or a robot pick from a tray. If your roadmap includes a closure like that, put it on the requirement list at concept stage. Retrofitting a second feeding technology onto a finished machine is a rebuild.
Two mechanical points that decide uptime more than feed rate does. Cap tracks need to be cleared and changed without tools, because they will jam and an operator will be standing there. And the pick-off point needs a sensor confirming a cap is actually there, so the machine skips the container instead of capping air and creating an open bottle downstream.
What Actually Goes Wrong on a Capping Station?
Capping failures are a short and repeatable list. A machine is worth buying when it catches every item on it, and each check costs money, so it is worth knowing what you are paying for.
| Failure mode | What it looks like on the floor | Usual root cause | What catches it |
|---|---|---|---|
| Cross-threading | Cap sits high, turns without tightening | Cap presented off-axis, worn chuck jaws, container not held square | Torque curve rises immediately with no run-down, plus cap height out of band |
| Cocked or skewed cap | Visible tilt, leaks under compression | Cap dropped off-centre from the chute, container not centred in the starwheel | Top-view camera measuring skew angle |
| Torque drift across a shift | Same recipe, falling removal torque at QC | Chuck or clutch wear, liner batch change, thread lubricity change | SPC trend on the logged torque data, not a single reading |
| Cap missing or not seated | Open container, spilled product downstream | Feeder starved, escapement jam, cap bounced out at pick-off | Keyence IL-600 laser displacement measuring cap height, and a no-cap-no-run interlock |
| Tamper band already broken | Pilfer ring separated before dispatch | Application torque too high, band bridges weak, chuck gripping the band | Top-view camera checking pilfer ring separation and tamper band engagement |
| Liner missing or displaced | Passes torque check, leaks in transit | Closure supplier batch problem | Not caught by a standard capping station; needs a dedicated cap-inspection camera before application, specified at design stage |
| ROPP thread not fully formed | Cap spins freely on a filled bottle | Roller pressure or height set wrong, shell thickness variation, glass finish out of tolerance | Formed-thread inspection plus offline removal-torque testing on samples |
| Cap backs off after release | Torque logged as correct, cap loose hours later | Liner elastic recovery, dwell too short at high speed | QC removal-torque testing on aged product, not by the machine |
The last two rows are the honest part of the table. A capping machine cannot see everything. Removal torque on aged product is a laboratory measurement, and no inline sensor substitutes for it. What the machine can do is hold the application window tightly enough that the laboratory result stays inside its own specification, and give you the data to investigate when it does not.
How Is Every Cap Verified, Not Sampled?
Every container on a Motionwell filling and sealing line is verified rather than sampled, and closure verification runs three independent checks that fail in different ways.
The laser displacement sensor measures cap height. It is a direct geometric measurement of whether a cap is present and how far down it went, and it is fast enough to run at full line rate. The torque curve covers the application itself, as set out above. Then a top-view camera inspects the sealed closure for tamper band engagement, pilfer ring separation, cap presence, cap skew angle and visible defects such as cross-threading or incomplete sealing, checking multiple criteria at once and assigning one composite pass or fail result.
Failed containers are diverted to a quarantine bin by a pneumatic pusher downstream of the vision station, and each rejection is logged with container serial number, rejection reason and inspection image for quality review and SPC analysis. How the cameras are specified, lit and calibrated, and how the false-reject and false-accept balance gets set, is on our machine vision inspection capability page.
State the limit clearly. None of these three checks measures seal integrity. Vision cannot see through material and cannot detect a leak; leak detection needs pressure decay or tracer gas, which is its own station with its own cycle time. If your product needs proof of container closure integrity rather than proof of correct application, that is a separate machine on the requirement list, and it should appear in the specification from the start rather than as a change order after commissioning.
When Should the Capper Be a Separate Machine Instead of Part of the Filler?
This is the question that gets decided too late on most projects, so here is the reasoning we apply.
Keep capping inside the filling machine when one container envelope covers your whole range, the closure family does not change, and the two stations naturally share a rate. Integration is cheaper, the footprint is smaller, one control system handles the handshake, and there is no transfer between machines to lose containers in.
Split the capper out when any of these is true.
The format envelope is different. On a rotary machine the star wheel, container guides and capping chuck geometry are all tied to a container envelope. A container outside that envelope is a change part set, not a recipe. If your bottles share a fill volume but not a neck finish, a separate capper with its own change parts costs less over five years than forcing both through one carousel.
Closure physics differ, not just size. A line running screw caps and trigger sprays is running two feeding technologies and two application methods. That is two machines wearing one cabinet, and it will be quoted and maintained as such whether or not it is drawn that way.
Changeover has to happen while the filler keeps running. A standalone capper can be set up on the bench for the next SKU and swapped in. A capping section inside a rotary filler stops the filler.
You already own a filler that works. The most common real case. If dosing is fine and closing is the problem, buying a filler to fix a capper is an expensive way round. A retrofit capping station with its own conveyor tie-in is the direct fix.
Validation scope should be contained. In a GMP plant, changing anything inside a qualified filling machine drags requalification across the whole machine. A separate capping station bounds that scope.
The rates do not match. The pharmaceutical platform runs 120 bottles per minute at a 100 mL fill from eight heads. If the closure needs a longer dwell than the fill does, the capper needs more heads or its own carousel, and buffering between two machines is easier than balancing inside one.
Against all of this, integration wins on footprint, on cost, and on one less container transfer. Two machines mean an accumulation conveyor between them, and every transfer is a place a container falls over.
What Changes When the Capping Line Is GMP?
The mechanism does not change. Everything around it does.
Product-contact and splash-zone construction goes to SUS316L electropolished to Ra 0.4 um, welds ground flush and passivated to ASME BPE practice, with gaskets and O-rings compliant to FDA 21 CFR 177.2600 and clean-in-place spray ball ports at drain points. The food-grade platform runs SUS304 and 316 at Ra 0.8 um with IP65 enclosures for daily washdown, which is a different discipline aimed at a different risk. Both are covered on the pharmaceutical packaging automation page and the food and beverage automation page.
Documentation becomes half the project. GMP builds ship with IQ/OQ/PQ packages, factory acceptance testing at our Singapore facility before shipment, site acceptance testing at your cleanroom, and validation protocols following ISPE Baseline Guide Vol. 5. Torque records support 21 CFR Part 11 batch records.
One interaction catches projects out. Where the pack is serialised, the closure and the code compete for space. The line prints and verifies 2D DataMatrix codes with Cognex DataMan readers and Domino thermal inkjet printers at up to 150 units per minute for Singapore HSA and EU FMD compliance, and the mistake we see most often is a code printed on a panel that later ends up under a shrink band, a tamper seal or a cap skirt. Walk the pack through every downstream station with a marker pen on the intended code position before anyone orders equipment. The coding side is on our pharmaceutical serialization page.
Can a Capping Station Be Retrofitted to a Line You Already Run?
Yes, and it is a large part of what we do. Control system modernisation of existing production machines, replacing ageing PLCs, servo drives and VFDs across Allen-Bradley, Siemens, Omron, Mitsubishi and Beckhoff platforms, is our largest line of work this year. A retrofit capping station is the same problem with a mechanism attached.
Three things have to line up. Mechanically, the new station has to accept containers at the height and pitch the existing conveyor runs at, and hold them square against the reaction torque, which is often the harder half. Electrically, it needs a handshake with the line: container present, cap available, station ready, reject asserted, fault. And the reject has to go somewhere, which means a quarantine bin with a full sensor rather than a chute onto the floor.
Two delivered patterns are worth knowing about. The vial capping and uncapping station on P23005 uses a collaborative robot with a custom end-of-arm tool and pneumatic torque control with cap presence verification, which suits laboratory and low-rate GMP work where the container arrives in a rack rather than on a conveyor. And where floor space is the binding constraint, we mount JAKA collaborative arms overhead for pick-and-place and capping, which frees the floor area for conveyors and operator access at the cost of harder maintenance access. That is a trade, not a free upgrade. Safety scope for either follows from a risk assessment, as set out on our machine safety and compliance page.
What Drives the Cost and Lead Time of a Capping System?
We do not publish prices for a capping machine in Singapore, because the same nominal machine moves by a wide margin on decisions taken before hardware is ordered. What we can be specific about is which decisions move it.
| Cost driver | Why it moves the number |
|---|---|
| Closure count and type | Each closure family can bring its own feeder, chuck and change parts |
| Head count | Set by rate and by dwell time, and each head carries a servo, a chuck and a calibration record |
| Container format range | Star wheel, guides and chuck geometry are tied to an envelope; more envelopes means more change parts |
| Feeding difficulty | A symmetrical screw cap sorts cheaply; a trigger spray or hinged flip-top does not |
| Verification depth | Cap height alone, versus height plus torque logging plus a vision station plus reject handling |
| Data and validation scope | Torque records as Part 11 evidence is a documentation project alongside the build |
| Construction standard | SUS316L at Ra 0.4 um with CIP is a different build from a food-grade washdown frame |
| Integration scope | Dry contacts to an existing line, versus recipe download from MES and full line control |
Lead time runs 16 to 24 weeks from concept approval to factory acceptance testing on a standard build, and 24 to 32 weeks where cleanroom compatibility or full GMP validation applies. Design, fabrication, assembly and testing happen at Woodlands Link, which is why a Singapore buyer attends the factory acceptance test instead of flying to it. How a project of this shape is scoped from concept to commissioning is in our write-up on special purpose machine design.
What Will We Not Take On?
Worth being direct, so nobody spends a month finding out.
We do not manufacture caps, closures or liners, and we do not select them for you. We do not issue CE certificates and we are not a notified body; we build to a specification and support your conformity work, including LVD and CE testing and MOM lifting certification where the machine includes lifting equipment. We do not build production welding cells or sell software products. We have not designed an induction sealing head and we integrate a proven one instead. And where your closure runs at a rate a standard catalogue capper already covers at a lower price, the useful answer is to say so, which costs you a conversation rather than a commitment.
For container sizes above the bottle and vial range, closure handling changes completely: pail lids are pressed and clinched rather than torqued, and drum bungs are located before they are capped. That work is on the drum and pail filling page. Downstream of closing, the line continues into coding, case packing and automated palletizing.
Frequently Asked Questions
What capping torque can Motionwell hold, and is it recorded?
Programmable from 0.5 to 5.0 Nm, with the torque curve logged for every container rather than sampled. The rotary capping heads on projects P23005 and P25026 run Mitsubishi HG-KR servo motors with torque feedback, so the controller reads the whole application curve instead of a final number. The target torque itself comes from your closure supplier and your own removal-torque testing. We build the station to hold it, prove it at factory acceptance testing, and keep the record.
Should the capper be part of the filling machine or a separate station?
Keep it on the filler when one container envelope covers your range and the two stations share a rate. Split it out when closure formats differ in physics rather than in size, when the capper needs its own changeover while the filler keeps running, or when you already own a filler that works and only the closing station is the problem. On a rotary machine the star wheel, container guides and capping chuck geometry are tied to one container envelope, so a container outside it is a change part set, not a recipe.
How do you verify that every cap is applied correctly?
Three independent checks, applied to every container rather than to a sample. A Keyence IL-600 laser displacement sensor measures cap height to confirm the cap is present and seated. The servo torque curve confirms the application itself. A top-view camera checks cap skew, tamper band engagement, pilfer ring separation and visible cross-threading, and assigns one composite pass or fail. Failures go to a quarantine bin by pneumatic pusher, logged with serial number, reason and image.