Biometric sensor
Development of the casing and mechanical design of a near-medical device
Embedded Files
Once upon a time - no one really remembers exactly when - there lived a client. And this client had a device. Its looks and functionality had both grown old and tired - technically outdated, and aesthetically too. Not a total disaster, of course, but let’s be honest: with that kind of gadget, you wouldn’t exactly set sail to conquer new markets overseas. Even at home, it had become hard to compete. Pure specs don’t win hearts anymore - people want good design these days
So the client decided to make a new product — one whose housing would support the growing needs of the electronics inside, and one that could sell on looks alone
Well, let’s see what we can come up with
Since the new PCB hadn’t yet been developed at the industrial design stage, the internals from the previous device were used as a base. All sketching and modeling revolved around that setup. Naturally, once the project reached the mechanical design phase and we started laying out the new components together with the client’s electronics engineer, it quickly became clear that there was simply no way to fit everything into the existing dimension
It’s worth noting that arranging components inside an electronic device isn’t like packing a suitcase with all sorts of stuff before a vacation. Components follow certain layout rules, and very often it turns out that while there seems to be plenty of “empty” space, in reality, there’s nowhere suitable to place anything. So, right off the bat, the PCB grew from 51×23 mm to 55×24 mm. In absolute terms, that’s nothing - but for such a small device, it’s quite noticeable
After a round of discussions and some reflection, the electronics engineer came back with something almost straight out of an old joke - start worrying, details to follow - saying: “…the electronics don’t fit, and by a lot… I’ll send my questions and suggestions later… for now I’m increasing the board width to 26 mm”. Which, of course, led to the overall device width increasing by another 4 mm
We then “seated” the PCB, adjusted the position of the USB connector so that the plug would sit snugly against the housing (which is only possible if you’re using one specific connector type -- which was exactly the client’s case)
The ind. designer double-checked the model for surface aesthetics, and the client went on to order the PCB and 3D-print a prototype to verify assembly
In this project, the client handled all production themselves -- not the most typical setup, but perfectly acceptable as long as feedback is provided promptly so the design side stays aligned with any critical updates. When the first version of the board was installed into the photopolymer-printed housing, one key issue emerged: the PCB would need to be re-routed -- moving more components to the top side, reserving extra space below for future system expansion, replacing the radio module, and so on
Technologies: multi-shot injection molding (overmolding)
Materials: ABS, TPE, PMMA, PC
Planned production volume: 30000 units
Operating conditions: indoor, IP50, from -40 to 40°C,
Protection against: impact at the IK03 level
My role: design engineer
Time spent on the project: 153 h.
Project duration: 14 mo.
After another electronics–housing iteration, the client moved on to prototyping. It was decided to prototype using two methods: vaccum casting and CNC machining, with the silicone-cast "rubber" components later glued onto the machined parts. In terms of differences, the first method (VC) produced better-looking results - especially in this case, with two components, where the process could roughly reproduce a series-production setup - while the second method (machining) yielded higher dimensional accuracy and more representative mechanical properties compared to the future production parts. Naturally, the prototypes revealed some issues. The main ones were:
1) Accidental opening on impact. The housing opened in about one out of five drops from a height of 1 m and in roughly one out of three drops from 2 m. The client, of course, wanted zero openings in 100% of cases. To guarantee that level of reliability with only a mechanical joint (latches), the shell would have to be rigid enough to prevent deformation that could disengage the latches during impact. In theory, this could be achieved, for example, by adding an internal metal frame to resist impact forces, by switching to a stiffer polymer, or by increasing the damping layer so that most of the deformation occurred within it. All these options, however, would not guarantee 100% success and would noticeably compromise the overall product. The only truly reliable options would be bonding or welding. Therefore, the lesser evil was chosen - applying adhesive to the latches during assembly.
2) Insufficient rigidity of the housing button. The button plunger moved too loosely and sometimes even slipped out of alignment with the tactile switch underneath. If this had been made of production-grade thermoplastic elastomer, the situation would probably have been better, though still not ideal. As a solution, I moved both the load-bearing and actuation functions into the first (plastic) component, while leaving the second one with only decorative and protective roles
3) Minor adjustments. Some small tweaks here and there - for example, reinforcing and lengthening the anti-loss feature of the USB cap. At first glance, this seemed trivial: just increase the element’s dimensions. However, a larger retaining feature required more internal space, which was already fully occupied by components. Through a series of iterative adjustments, a workable compromise was eventually found
What's this?
The device being developed is a sensor for electromyography -- that is, for measuring muscle activity
It attaches to the body, records the bioelectrical activity of muscles, transmits data via Bluetooth, and displays it in an app or on a computer
It allows users to see in real time which muscles are active and how intensely -- useful for fitness, technique correction, and athletic training
It’s not a medical device, but rather a fitness or sports gadget
Main users:
-- trainers,
-- advanced athletes,
-- rehabilitation sessions where it’s important to analyze muscle load, control asymmetry, and track performance efficiency
The client moved on to another round of prototyping. The result fully met his expectations, so it was time to move toward serial production. He also decided to handle that phase independently. Therefore, he was provided with all the source files (models and drawings), and our already not-too-intensive communication naturally shifted to occasional consultations on secondary matters -- for instance, how to communicate specific texture requirements to the manufacturer. It may not sound great, but that’s actually a good sign -- it means there were no problems
Summary (for the entire project, not just the engineering part I’m describing here): in just over a year, the client received serial production units of the new product. A solid result, in my opinion. If it seems like the whole process could’ve been wrapped up in a couple of months - well, that’s definitely not the case. In theory, with very (VERY) strong motivation on the client’s side, the timeline could’ve been cut roughly in half. But I’ve never encountered such a case myself
Summary of the work done: based on the designer’s concept, I developed a device structure optimized for mass production and aligned with the requirements of the electronic components
A bit of statistics: a day-by-day timeline of the entire project, with an approximate breakdown by type of work
The infographic presented here is purely statistical. It reflects how this specific project unfolded for a particular client, with certain contributors, during a defined period. With different initial conditions, the results might have varied
Actual project budget: $50,000 ± 5k (from project start to the first units hitting the shelves)
Page updated
Google Sites
Report abuse