Robotics and automation

Why this sector suits printed parts

You can buy the robot, the gripper body and the controller. What you cannot buy is the part that touches your product, because it has to match a shape only you handle.

That leaves every integrator making one-offs: jaw faces shaped to a specific component, a mount that puts a camera exactly where the vision system needs it, a fixture that holds a part still while the robot works on it, a guard that fits around a cell nobody has built before. These are low-quantity, application-specific, and frequently revised — which is the profile printing handles better than any other process.

They are also usually needed now. A cell that is commissioned and waiting on jaws is an expensive thing to have standing still.

What gets made

• Gripper jaws and fingertips shaped to the part being handled, including soft faces in flexible TPU where the product must not be marked. See cobot gripper jaws and EOAT for the detail.
• End-of-arm tooling — multi-part pickup heads, vacuum cup mounts, tool adaptors and changers.
• Sensor, camera and light mounts, where the position matters more than the part does and adjustment usually happens on site.
• Cell fixtures and nests that hold a part in a known position so the robot can find it repeatably.
• Cable management, covers and light guards — the parts nobody budgets for and every cell needs.
• Spacers, adaptors and shims that reconcile two components never designed to bolt together.

How a job usually runs

First article, then the rest. One is printed and tried on the actual robot with the actual product, because a jaw that grips reliably in CAD is a different thing from one that grips reliably at speed with a part that arrives slightly rotated. What is learned goes back into the model, and only then is a set printed.

Send the part being handled if you can, or a good scan or drawing of it. The geometry of the product decides the jaw, so it is the single most useful thing to provide. Also say what the cycle time is and how the part presents itself — those change the design more than most people expect.

Reorders are straightforward: the model is kept and versioned, so a replacement set is a print rather than a new job. Jaws are consumable by nature, and this is a good thing to have arranged before a set wears out rather than after.

The honest limits

Payload and speed have a ceiling. Printed tooling suits light and moderate payloads at ordinary cobot speeds. High-speed industrial arms slinging heavy parts impose loads and accelerations that want machined aluminium, and you will be told when a job is in that territory.

Nothing safety-rated. Printed parts do not belong in a safety function. Guards that are part of a risk assessment, interlocks, light-curtain mounts whose position is safety-relevant, or anything a compliance obligation attaches to should be made another way. This is not negotiable and it is better said now than during a commissioning audit.

Heat and chemicals. PLA, PETG and flexible TPU are the materials here. Cells running hot, welding cells, and anything with solvent exposure are outside what these materials should be asked to do.

Wear is real, and often acceptable. A printed jaw face wears faster than a metal one. Where the face is consumable and can be reprinted cheaply from a kept model, that is frequently the better arrangement — particularly for a soft face that is protecting the product. Where a jaw has to hold its geometry for years of continuous cycling, it is not.

Who this suits

Integrators, machine builders, automation engineers and manufacturers running their own cells. The pattern that works best is a first article, then a set, then reorders as they wear — and it is welcome as an ongoing arrangement rather than a queue of separate jobs.

Frequently asked questions

Do you make jaws for Robotiq, OnRobot and Universal Robots?

Yes — those are the platforms most commonly asked about. What matters more than the brand is the mounting interface and the part being handled, so send the details of both. The dedicated cobot grippers page covers the design considerations in more depth.

Can you work from the part we are handling rather than a drawing?

That is usually the better way. The product geometry drives the jaw, so the physical part, a scan or good photographs with measurements all work. Tell us how the part presents itself and what the cycle time is, because both change the design.

How fast can we get a first article?

Quoting is normally same-day and a first article is typically quick, because it is one part rather than a set. Where a cell is commissioned and waiting, say so at the enquiry and priority scheduling can often be arranged.

Can printed tooling handle a production duty cycle?

For light and moderate payloads at cobot speeds, generally yes, with the expectation that jaw faces are consumable and get reprinted. For high-speed high-payload industrial arms, no — that is machined-aluminium territory and you will be told so rather than sold a set that fails.

Do we get the model?

Yes, where you have paid for the design. That matters in automation, where a cell may be recommissioned years later by somebody else. See the file policy on the about page.