Skip to content
SMG SMGTEXYarn Feeding Technology Talk to an Engineer

OEM / ODM

Custom work starts with an honest answer.

Some requirements are a bounded change to a part we already make. Others are a new electronic product with its own hardware, firmware and test plan. We establish which one yours is before any price is discussed — and we say when a project is not a fit.

  • Two depths Part-level adaptation and electronic product development
  • Feasibility first Reviewed before a price is quoted
  • Sample before volume Nothing scales on an unverified design
  • Scope in writing Deliverables and boundaries agreed before design work

Two depths of customization

A part adjustment and a new electronic product are not the same project.

They use different engineering time, different tooling decisions and different test plans. The first question in every inquiry is which of the two is being asked for — because the answer changes the whole route.

Depth 01 / Part level

Adaptation of an existing part

Dimensions, thread or fit, material, surface treatment, a mounting that has to match your machine, or a variant of a part already in our range. The change is bounded by the existing design, so tooling and validation stay proportionate.

  • Dimensions
  • Material
  • Fit and mounting
  • Variant
See the project stages

Depth 02 / Product level

Electronic product development

A new feeder, a control or detection function: hardware, firmware, mechanical integration and the test plan that proves it on a real machine. Longer, reviewed as a full project, and not offered on every request.

  • Hardware
  • Firmware
  • Integration
  • Validation
See what it involves

Two depths of customization

Same stages, different depth of engineering.

A bounded change to an existing part and a new electronic product use different engineering time, different tooling decisions and different test plans.

The project route

From requirement to volume, in eight stages.

Both depths follow the same sequence; what changes is how much engineering each stage holds, and how early a project that does not add up can be stopped.

  1. 01 Requirement collection

    Machine type and model, yarn, application, quantities, and the problem the change is meant to solve. Photos or drawings of the current part answer more than a written description.

  2. 02 Feasibility assessment

    Whether it fits the machines and the process, whether it can be made repeatable, and whether it conflicts with an existing agreement. A no is given here, deliberately, rather than after tooling.

  3. 03 Scope and commercial terms

    What is delivered, what is excluded, who owns what, and the quantity expectation — recorded in writing before any design work starts.

  4. 04 Design

    Mechanical structure and, where electronic, hardware and firmware. Drawings and specifications are shared for your comment before anything is manufactured.

  5. 05 Sample

    A sample, or a small first batch, built to the agreed specification — your configuration, not a catalogue item with a different label.

  6. 06 Testing

    Function, run-in and, where the application requires it, validation on the machine — under the same control chain as standard supply.

  7. 07 Confirmation

    Results are reviewed together, points that do not meet the requirement are corrected, and the sample configuration is signed off as the production reference.

  8. 08 Volume production

    Production runs to the confirmed reference, with the same inspection, batch record and traceability as standard supply.

Image slot 1600 x 900: sample assembly of a first article at the workbench

Product level, in detail

What deep electronic customization involves.

Developing an electronic feeder means hardware, firmware, mechanics and test planning at the same time. It is a longer project than a part adaptation, and it is reviewed and priced as one.

  1. 01 Hardware

    Board, enclosure and wiring designed around the real working conditions: temperature, vibration, power supply and duty cycle.

  2. 02 Firmware and control

    Tension, speed and detection behaviour defined, with parameters the mill can set, repeat and record on its own machines.

  3. 03 Mechanical integration

    Mounting, yarn path and service access designed for the machine the unit goes on, not for a catalogue photograph.

  4. 04 Test plan

    The measured values that decide whether the design works: function, run-in, ageing and validation on the machine it was made for.

  5. 05 Documentation

    Drawings, settings and the confirmed reference handed over with the production batch, so the configuration is not lost when a unit needs replacing.

Every unit released under a custom project passes the same control chain as standard supply. See Quality & Testing

Image slot 1200 x 800: electronic feeder prototype on a test bench with instruments

Sample before volume

Nothing scales on an unverified design.

A sample, or a small first batch, is built to the agreed specification, tested under the same control chain as standard supply, and signed off as the production reference before volume runs.

Image slot 800 x 600: engineering drawing review with the part on the bench

Boundaries and intellectual property

What is agreed before design work starts.

Custom projects go wrong on assumptions never said out loud: who owns the tooling and the design, what happens if the project stops. Those are settled at the scope stage, in writing, while they are cheap to discuss.

  • Ownership Tooling, drawings and design ownership are named in the project agreement: what stays with the customer, and what stays with SMG.
  • Background technology SMG keeps the rights to the designs and production know-how that a project builds on. A design supplied by the customer stays with the customer.
  • Confidentiality Customer drawings and project data are treated as confidential, and a confidentiality agreement can be put in place before drawings are shared.
  • Exclusivity and sole supply Exclusivity, territory or sole-supply terms are raised at the scope stage and are only valid once they are written into the agreement.

The wording here describes how projects are run. It is not a quotation and not the agreement itself — the signed document, reviewed with you, takes precedence.

Feasibility, honestly

When the answer is no.

Not every inquiry becomes a project, and pretending otherwise uses the engineering time another customer is waiting for. Four cases where the assessment usually ends in a no.

  • It cannot be made repeatable A design that depends on one operator or one fixture is not something we can supply in the same state next year.
  • It conflicts with an existing commitment An existing agreement, an existing customer or an existing supply position can rule a project out before any design work.
  • The volume does not carry it Tooling and engineering have to be justified by the quantities involved. Where they are not, a standard product is the better answer.
  • It sits outside our process Knitting machine components and the electronics around yarn feeding. Work outside that scope is declined rather than taken on badly.

A no at the feasibility stage costs a conversation. A no discovered after tooling costs a project — which is why the assessment comes before the quotation, not after it.

Next step

Send the requirement and the machine.

Machine type and model, yarn, application, quantity, and what the current part or product cannot do. An engineer will come back with the depth of customization the job actually needs — and, if it is not a project we should take, the reason for that.