A prototype that clears its first bench run rarely clears it twice in the same way. The wiring changes, a bracket is re-machined, a controller setting is retuned, and the next unit is on the test rig before the previous result has been written down. That is the situation in which cycle time stops being a line in a quotation and becomes the thing that decides whether a prototype programme reports a result this quarter or next. Estimating it properly means working out what has to happen on the rig, what can happen while the rig is idle, and what is being counted as one cycle in the first place.
Decide what one cycle has to prove
Before any timing is estimated, write down the pass conditions for the unit under test: which parameters are measured, what tolerance each has, and which measurements must be repeated after a change rather than only once. A cycle counted from power-on to a green lamp is not comparable with one counted from fixture load to a signed report, and suppliers quoting against different definitions end up bidding different jobs. Put the cycle on paper as a sequence of steps with a named owner for each, and the estimate that follows will hold up when a programme slips.
Break the cycle into the four blocks that consume time
Most hardware test cycles fall apart into four blocks, and each responds to a different lever.
- Load and set-up. Fittings, cable harnesses, tooling and software loading. Repeat work here scales directly with programme size, so a fixture or a standard harness pays for itself quickly.
- Soak and stabilise. Warm-up, thermal hold, run-in. It is largely fixed and cannot be compressed, which makes it the right block to overlap with everything else.
- Measurement. The instrumented part of the run. Sampling, logging and switching between channels are what usually make this longer than the physics requires.
- Disposition. The decision after the run: pass, retest, or stop and investigate. An undefined pass threshold turns a short measurement into a long argument.
Count what happens between cycles
A rig that runs for two hours and then sits for half a day does not have a two-hour cycle. Changeover between prototype revisions is where this is lost, and it is a design problem as much as a scheduling one. PowerDe lists hardware testing among its engineering services alongside equipment manufacturing automation services, and the two sit together for exactly this reason: a test station built for a prototype that will be revised three times needs interfaces, fixtures and programs that can be swapped without rebuilding the cell. Ask the supplier how a change of revision is absorbed, and how long that absorption takes on their own line.
Use data to set the target, not ambition
The first cycle of a new rig is not a representative cycle; it is the slowest one you will ever see, because it includes commissioning, operator learning and the correction of a fixture that does not quite fit. Take the median of a dozen consecutive runs rather than the first one, and record the reasons for the outliers separately, since a retest caused by a bad connection is a reliability finding, not a throughput finding. Once the median and the spread are both known, the target stops being a wish: you can state what has to change — an added fixture, a second channel, an automatic log — and what that change is worth.
Compressing the cycle on equipment prototypes
Where cleaning or surface preparation sits inside the test sequence, it usually becomes the bottleneck because it is manual and done at a bench. Station-mounted equipment removes the walking and the refilling: the PowerDe HW02 steam cleaner is a 220 V fixed unit with 6 kW of heating power, 160℃C steam at 0.8 MPa and an 80 L SUS304 tank, fed from the plant water supply, and it switches between dry and wet steam so that a preparer can work wet where residue is heavy and dry where the part must stay nearly dry before it is instrumented. The same logic applies to any fixed process step at a test station: the time saved is mostly movement and waiting, and that is why placement matters more than the specification sheet.
Where an external team shortens the path
A small programme rarely justifies hiring a rig engineer for one instrumented cell. Where the internal team is thin, a supplier that takes the test requirement and returns a working station — fixtures, programs, documentation and the hand-over — moves the cycle-time question from your schedule to theirs. The engineering services listed at https://www.powerde.net/gongchengshi/ cover this kind of scoped work, and enquiries about a specific prototype go through the contact page at https://www.powerde.net/contact/. Send the drawing, the pass criteria and the volume you expect per week; those inputs are what let a supplier say what the cycle will cost in minutes.
