Start with the line, not with the machine list
A custom production line is not a set of machines with a conveyor between them; it is one process with one takt time. So the first job when you choose an engineering partner is to describe the process you actually need: what goes in, what comes out, the rate you have to hold, and which steps are fixed by your product and which are open to change. A partner who receives a machine list will build the list. A partner who receives the process will tell you which machines the process needs, including the ones you did not consider.
Check that the partner builds what you make
Ask for evidence in your own process area, not in adjacent ones. Custom-line work is unforgiving: the assumptions that matter are specific to the material, the tolerance and the environment. PowerDe, for example, works across intelligent cleaning equipment, semiconductor process equipment and thin-film solar equipment, and its line-level work includes automated production solutions such as its Industry 4.0 demonstration factory listing. When a partner has already built in your process family, the questions you get back are sharper — and sharper questions before a contract are worth more than a lower price with generic answers.
Test the partner's engineering depth with three questions
Ask what they would change about your current process, what they would not automate and why, and what their last custom line did badly. A partner worth hiring answers all three concretely. A partner who only answers the first is quoting, not engineering. It is also worth asking who at the company will own your project end to end; some partners offer an integrated engineering team that carries development, integration and commissioning under one thread, which is what keeps a custom line's interfaces somebody's responsibility.
Make them define the handover before you sign
Custom lines are usually lost at handover. Ask, in writing, what the delivered package contains: the as-built drawings and revision records, wiring and utility diagrams, the parameter and recipe list, the supplier's own test results, the spares and consumables list, and the operating instructions in the language your operators read. Then ask how acceptance is tested — what has to run, at what rate, for how long, and who signs it off. A partner who already has this structure has delivered lines before; a partner who finds it awkward after signing will be improvising on your budget.
Check what happens after the line starts
The support terms are part of the selection. Agree response time, remote or on-site support, who holds the consumables and the critical spares, and how a fault is attributed between your team and theirs. Also agree what happens when your product changes: on a custom line, a product revision can mean a fixture change, a recipe change or a station re-work, and the price of that change should be discussed while you still have leverage.
Run a paid stage before the whole line is committed
Do not let a custom line be one contract from drawing to acceptance. Structure it so the first paid stage is the process study and the layout, and the second is the trial of the critical station on your own product. That converts the largest risk — "the hard station will not reach the rate" — into a small, early, checkable question. A partner who will not accept a staged structure is asking you to carry all the risk.
What to send when you start the conversation
Send the product family, your current cycle time, the steps you cannot change, the utilities you can supply, and the rate you need in twelve months rather than today. Send it to PowerDe's engineering team through the contact page and ask for the layout critique and the trial proposal, not a quotation. The quality of that first response is itself the best selection test you will get.
