Price the phone first.
The idea arrives often: we need a device that does this one job, so let us build one. Sometimes that is right. More often, a plain phone or tablet already has a screen, a battery, a microphone, a camera, radios and a software store, all tested by millions of users.
A custom device takes on every one of those parts as your problem. That includes boot time, battery life, firmware updates, radio approvals and spare supply. It is worth taking on only for something the phone clearly cannot do.
So test the phone first, and write down what it fails. That list is the real specification for anything custom.
What to do
- Write the tests the device must pass in plain terms: boot time, battery life over a shift, drop onto concrete, microphone pickup in noise, readability in sun.
- Run a plain off-the-shelf phone or tablet through every test and record the result.
- Mark each test pass or fail. Do not argue with the results.
- For each fail, ask whether a case, a mount, a setting or an accessory fixes it.
- Whatever is left, such as a physical mute switch you can trust by touch, is what justifies custom hardware.
- Scope the custom device around only that remaining list.
When not to
If the phone passes everything, build nothing. That is a good result. If your use needs something the phone cannot offer by design, such as an unusual sensor, a strict safety behavior or a tough environment, skip straight to prototype work but still keep the test list.