INTERVIEW | 4 min read

The Hidden Work Between 'It Works Once' and 'It Works Every Time': An Interview on Custom Hardware and Embedded Engineering

Last edited: Nov 16, 2026
Listen
--:--
The Hidden Work Between 'It Works Once' and 'It Works Every Time': An Interview on Custom Hardware and Embedded Engineering hero image

The Hidden Work Between 'It Works Once' and 'It Works Every Time'

Custom hardware and embedded engineering projects rarely fail because the board is too hard to design. More often, they slip because of the work that happens after a prototype first powers on. In a short interview, ElecOrigin, a firm working across hardware, embedded systems, and software development, described where timelines actually go and what general guides tend to omit.

Where the schedule really goes

Asked what takes far longer than customers expect when they arrive with a product idea and a deadline, ElecOrigin pointed to the gap between a demo and a dependable product.

They said "the part customers underestimate is everything between 'it works once' and 'it works every time': clarifying what 'done' means, testing, and revising the design and firmware together." The board itself, they added, is rarely the slow part. "The rework loops are."

That framing is consistent with broader embedded systems engineering practice, where verification and iteration often consume more calendar time than initial design. For teams planning a launch, the practical takeaway is to budget explicitly for test cycles and design revisions, not just for the first build.

Revision discipline: the part guides leave out

The second question asked what keeps coming up in custom hardware and firmware projects that online guides tend to skip. ElecOrigin's answer was specific: revision discipline.

They said "guides show a single clean build, but real projects live in revisions: which schematic revision is current, which firmware version was actually tested, and what counts as a pass." The consequence is concrete: "Most delays come from building or testing the wrong revision."

That is a process problem as much as a technical one. Teams that treat version control as something only for software, or that leave schematic and firmware revision tracking informal, tend to discover mismatches late. ElecOrigin's point suggests a simple discipline: make the current schematic revision, the tested firmware version, and the pass criteria visible to everyone on the project, and treat a mismatch as a stop signal rather than a detail.

The decisions that are expensive to undo

The third question asked what people decide too quickly at the start of a build, before talking it through with an engineer. ElecOrigin listed three recurring commitments.

They said "people lock in a timeline, a parts list, or a form factor before the brief is settled: what the product must do, who uses it, and what counts as tested." Those early choices, they noted, "are the most expensive to undo."

This is a familiar pattern in new product development: schedule and bill-of-materials decisions feel concrete and reassuring, while the brief feels abstract. But a parts list chosen before the requirements are clear can constrain the design, and a timeline set before testing scope is defined can create pressure to skip the very rework loops that determine whether the product works reliably.

Practical takeaways

Across the three answers, a consistent theme emerges: the expensive parts of custom hardware work are not the schematic or the first firmware build, but the coordination around them. Three practices follow directly from ElecOrigin's responses.

  • Define 'done' before the build starts. Clarify what counts as tested and what counts as a pass, so testing and revision have a target.
  • Track revisions explicitly. Know which schematic revision is current and which firmware version was actually tested. Treat mismatches as blockers.
  • Settle the brief before locking the timeline, parts list, or form factor. Requirements, users, and test criteria are cheaper to change early than late.

For teams commissioning custom hardware, these are unglamorous recommendations, but they map to where ElecOrigin says delays actually originate. The full scope of the firm's work is described on the ElecOrigin website.