Wearable Device Design: How a Standalone GPS Bracelet Was Built

Wearable device design is the practice of fitting electronics, power, antennas and a human-friendly enclosure into a volume that a person will willingly wear all day. Every requirement competes for the same few cubic centimetres, so the design is a sequence of explicit trade-offs between battery life, size, comfort and function. This is one build where those trade-offs are easy to see: a GPS bracelet that reports its location without a phone anywhere near it.

Most consumer trackers cheat. They pair to a phone over Bluetooth and let the phone do the locating and reporting. That is useless for the two groups this client cared about - young children and elderly relatives - because neither reliably carries a phone.

So the bracelet had to be self-contained: its own GPS receiver, its own connectivity to report position, its own power, and a form small and comfortable enough that a child would keep it on. It also had to detect falls and unusual motion, and survive being worn in the rain and the bath.

There is no arrangement that satisfies all six. The design work is deciding, deliberately and early, which ones win.

The hardest constraint was width. A wrist product for a child cannot be much more than an inch across and still be worn, and into that envelope we had to place a GPS module, the radio hardware needed to report position, the antennas both of them require, a battery large enough for a useful day, the sensors, and charging contacts.

Antennas are the unforgiving part. They need clearance and they hate being surrounded by metal or pressed against the body, both of which are guaranteed on a wrist. Battery placement, board outline and antenna keep-out zones all had to be resolved together, which is why a stock development board was never going to work.

On a standalone wearable, battery life is not a specification you choose at the end - it is the constraint everything else is designed around. A cellular radio transmitting a fix draws orders of magnitude more current than the device does at rest, so runtime is set by how often you report, not by how efficient the enclosure is. Decide the reporting interval first; the cell size, the thickness and the strap all follow from it.

Two radios in a wrist-sized enclosure means two antennas that must not detune each other, must not be shadowed by the battery, and must survive being pressed against a body of salt water - which is what a wrist is, electrically. Placement is decided with the mechanical layout, not after it.

Moving a battery two millimetres late in the project can cost more than the battery did, because the antenna has to be re-tuned and the housing re-cut. This is why we run enclosure and electronic design as one activity rather than two handoffs.

Anything with a radio has to be tested and approved before it can be sold, and lab slots are booked weeks ahead. Teams routinely finish a device and then discover that the market they planned to launch in needs three months of testing they did not budget. Scope this in the first month: which markets, which marks, which lab, and what a re-test costs if the first attempt fails.

Wearables age quickly: cells lose capacity, modules go end-of-life, and networks retire the bands you designed around. The cheapest insurance is to keep the board outline and mounting scheme stable so a module swap does not become a new enclosure, to choose a cell with a second qualified source, and to make firmware updatable in the field. Otherwise the second version is a second full development programme.

A connected device is not a one-off sale; it carries a monthly cost per unit in the field for as long as it works. Products in this category have died with good hardware and no viable operating model, because the connectivity and backend bill was never priced into the retail figure.

The instinct is to make something that looks like the product. The better first build is ugly and answers the risky question: does it get a fix indoors, does the radio hold a connection on a moving body, and how long does the cell actually last at your reporting interval. A shoebox-sized breadboard version on a strap tells you more in a week than a beautiful non-working model tells you in a month.

A device that reports a person's location, often a child's, carries obligations that shape the architecture. Retention periods, who can see a track, how an account is recovered and what happens when a device is resold are all decisions to make before the backend is written, because retrofitting them means migrating data you should not have kept. Treating this as a launch-time legal review is how products end up rebuilding.

Connected wearables generate more support contacts per unit than almost any other consumer product, and support cost can quietly exceed hardware margin. Most tickets cluster in a few areas that are fixable in design rather than in staffing.

We run this as a single mechanical and electronic design engagement rather than two handoffs. The cost breakdown shows how the stages are usually funded.

If the device reports to a person, yes - the app is the product for most users, and it usually costs as much as the hardware development. Plan it as a parallel workstream with its own budget rather than a final step, and keep the first version small: setup, live status, history and alerts. Everything else can wait for real usage data.

Schematic, layout, bring-up and firmware for a module-based connected wearable typically run $35k - $120k, before certification and tooling. A custom radio design rather than a module adds substantially and lengthens the schedule, which is why most first products use a module and revisit that decision at volume.

Twelve to twenty-four months to a certified, shippable product. Electronics and firmware run in parallel with mechanical design for six to nine months, integration and testing take another three, and certification plus tooling occupy the rest. Devices using pre-certified modules and an existing platform can compress the front half considerably.

Discovering the power budget late. Nearly every painful wearable programme we have seen went the same way: a form was agreed, electronics were fitted into it, and the resulting runtime was unacceptable - at which point the housing, the antennas and the tooling all have to change together. Measure real current draw at the intended duty cycle in the first month.

Outdoors with a clear sky, a few metres. In cities among tall buildings, tens of metres, and indoors often no fix at all without assistance from cellular or Wi-Fi positioning. Any product promising indoor accuracy needs a second positioning source designed in from the start, and the marketing has to be written against real measured performance rather than the chipset datasheet.

For the first version, usually yes, and it is often the right call - a certified module removes a chunk of radio testing and shortens the schedule by months. The trade is thickness and unit cost. Products that ship in volume typically start on a module and move to a custom design once the demand justifies the engineering, which is also the point at which the saving is real.

The battery sets the floor. For a device reporting every fifteen minutes on cellular, roughly 9 - 13 mm at the thickest point is normal once the cell, board, antennas and housing walls are stacked. Slimmer is achievable by reporting less often, by using a smaller radio, or by accepting a one-day runtime - all product decisions, not engineering tricks.

Choose a rating and design to it from the start, because sealing is architectural. Splash resistance can be achieved with gaskets and a well-designed parting line; full immersion generally means sealed charging - wireless or contactless - and a housing that is bonded rather than screwed. Retrofitting immersion resistance to a device designed for splashes usually means a new enclosure.

A connected wearable with a custom board, a moulded enclosure and firmware typically runs $60k - $200k to production-ready, before tooling and certification. Radio certification alone is usually $10k - $30k depending on the markets you enter. The cost breakdown by stage shows where that money goes.

Neither, in sequence. Wearables are the clearest case where the board outline, battery, antennas and housing are drawn together in one model from the first week. Designing a beautiful housing and then asking an electronics team to fill it is the most reliable way to produce a device that is either too thick or does not connect.

Wrist-width devices under an inch across are achievable with a custom board and careful antenna placement, but the battery becomes the limit rather than the electronics. Below that envelope you are usually trading run time hour for hour against millimetres, so decide the daily run time you must hit before you fix the industrial design.

For proving the concept, no - a development board and a rough housing answer the functional questions faster and cheaper. For a product that has to be worn, almost always yes, because off-the-shelf modules are rectangular and generously spaced, and that wasted volume is exactly what a wearable does not have.

Two routes. Either a sealed housing with gaskets and a sealed charging solution, which stays serviceable, or full encapsulation of the electronics in a material such as silicone, which is more robust and more comfortable but means the device cannot be opened. Wrist products worn by children usually justify encapsulation.

See the project write-up in the GPS tracking bracelet case study. If you are developing a connected product, see electronic design services and prototyping, or scope a build with the prototype cost calculator.