DTO part decoration — lessons earned at production speed.
The seven costly mistakes EPS engineers see manufacturers make when they implement direct-to-object inkjet — and how to avoid each one before commissioning.
What goes wrong, why it goes wrong, and what to do instead.
This paper is a postmortem in seven parts. Drawn from EPS field engineers who have commissioned hundreds of DTO inkjet cells, it walks through the recurring failure modes — undertested adhesion, mismatched automation, missing inspection, underspecified pretreatment, post-processing handling, color management without reference, and labor-handling assumptions that don’t survive multi-shift production.
For each, it explains what the symptom looks like, what the root cause usually is, and what to engineer differently next time.
There’s an old Henry Ford line we keep coming back to: “The only real mistake is one from which we learn nothing.” After many years of building direct-to-object decoration systems for customers who arrive with a specific part and a specific print in mind, we’ve learned some genuinely expensive lessons — and the most useful thing we can do with them is hand them to the next person, so the same mistakes don’t get made twice.
Most of those lessons come down to one idea: a direct-to-object inkjet system is not a printer you buy and point at a part. It’s a process you develop around the part. Before we quote anything, we work through a sequence of practical considerations — ink, print head, pre-treatment, curing, and the system around them. Here’s what that looks like, and why each decision matters more than it first appears.
Start with the ink, not the printer
The most common reason a DTO project disappoints is that the ink was chosen last. It should be chosen first. Before anything else, we need an ink that does three things on your specific substrate:
- It adheres — against a defined specification, not a hand-wave. “Sticks well” isn’t a spec; a cross-hatch and tape-test result is.
- It prints well — good wet-out and appearance, again against a described set of parameters for resolution and quality. An ink that adheres but beads or mottles isn’t a solution.
- It meets your requirements — UV-cured, FDA-compliant, food-grade, weatherable, chemical-resistant — whatever your application and regulatory environment demand.
Get this wrong and nothing downstream can save you. Get it right and the rest of the system has something solid to build on.
Choosing the print head
The print head determines what the system can actually produce, and it’s a balance of several factors:
- Print quality — resolution, ink lay-down, and the maximum and minimum drop sizes the head can deliver. Fine text and smooth gradients live or die here.
- Throughput speed — the question we always ask: what throughput do you need to pay for the machine in the time frame you’ve agreed to? Speed isn’t a vanity number; it’s the denominator of your ROI.
- Ink type — UV, solvent-, or water-based, which has to match both the ink decision above and the curing decision below.
- Image width — if your print needs, say, a 100 mm (4″) swath, a head wide enough to cover it in a single pass is far preferable to stitching two passes together. Stitching is a place where quality problems hide.
Pre-treatment is almost always part of the job
Given the variety of substrates in direct-to-object work — and especially the low-surface-energy plastics that dominate it — pre-treatment is almost always required to get the ink to bond. Three questions decide how it fits:
- What type works best and is compatible with the system you’re considering — corona, flame, plasma, or a chemistry like Pyrosil®?
- Will the pre-treat speed match the print speed? If it can’t keep up, you may need additional pre-treatment units or an accumulator to buffer the line.
- What are the risks? Some methods involve open flame; others involve high voltage. Ask the questions and brainstorm the failure modes before the equipment is built, not after.
Curing: can you harden the ink in the time you have?
Post-treatment — curing — is where throughput and chemistry meet. What kind of curing does the ink require: UV, hot air, a heated platen? If it’s UV, how much power do you need, and can you deliver it in the time the part spends under the lamp at your target line speed?
If the answer is no, you have options: increase the lamp wattage, add a second lamp or curing station, or introduce a post-print accumulator that effectively slows the part’s transit and raises the cure dose. The point is that curing capacity has to be designed against your real line speed — not assumed.
The considerations that are easy to forget
A handful of factors don’t fit neatly into ink / head / pre-treat / cure, but they sink projects just as reliably:
- Color gamut — can the system hit the colors your brand actually needs, including the ones outside a basic CMYK set?
- Primers and topcoats — sometimes a primer or a varnish topcoat is what gets you from “prints” to “passes the durability test.”
- Product handling and automation — how parts are fed, fixtured, and removed, plus any secondary operations like assembly or bagging. These shape the cell as much as the printer does.
- Environment — temperature and humidity on your actual production floor influence ink behavior and cure. A process validated in a lab can drift on a hot, humid afternoon.
How the pieces come together: process development
None of these decisions is made in isolation — we work them as a loop. We define the project requirements, then define the ink chemistry against adhesion and print quality; that informs the print-head selection (quality, speed, ink type, image size); which informs the pre-treatment (speed, cost, safety) and the curing (speed, size, power, form factor). Then we print samples, refine the process against the real results, and fold in the remaining requirements — colors and special effects, product handling, secondary operations. Each loop tightens the system around your part.
That’s the difference between buying a printer and engineering a decoration cell. The printer is one decision among many; the system is the answer. If you’d like to walk through these considerations against a specific part, bring us the part, the print, and the production numbers — the engineer who reviews them is the one who’d build your cell.
What we wish manufacturers asked us before they bought the printer.
Adhesion validated to spec.
Day-one prints can pass cross-hatch and crock tests and still fail in the field three weeks later. Real adhesion validation tests against the actual application environment — temperature, humidity, abrasion, chemistry — not just the print event.
Pretreatment scoped to the line.
An adhesion-grade pretreatment on the bench is meaningless if the production cell doesn't integrate it inline. The bench process is the prototype; the cell process is the product.
Inspection inside the cell.
Catching a misprint at the box is too late. Catching it inline lets the cell auto-reject before the part reaches packaging — the difference between 'good enough QC' and 'zero defects shipped.'
Color managed against physical reference.
Photographic brand color drifts. Without a physical reference panel locked under change control, brand managers reject prints that 'look fine to the operator.'
Automation upstream AND downstream.
Most lines automate the print station and underbudget the feed and unload. Operator labor migrates to the under-automated end and the savings collapse.
Variable data architected, not bolted on.
Adding lot codes and serialization after commissioning is twice the cost of building the data path in from the start. Plan the data layer before the print recipe.
Labor model that survives a shift change.
An operator-friendly cell on first shift becomes operator-hostile on third shift. The cell architecture has to account for the least-experienced operator who will ever run it.
The printer was almost never the problem. The problem was almost always something upstream or downstream of the printer that nobody scoped during the buying decision.
— EPS field engineering, twenty years of postmortems
Get the full paper.
The 7-page PDF — including the cost curve, the labor-savings table, and the break-even chart.