Smart Devices and Serum Absorption: What Sensing Buys You
Contact sensing, dose control and adaptive output — where they genuinely improve serum delivery, and where they are decoration.
March 14, 2026 · Updated April 7, 2026 · 4 min read
“Smart” is the most overused word in the beauty device category. For an absorption device, though, there is a real version of it: a device that knows how much serum it has delivered, and stops or adjusts accordingly.
This article separates the sensing that changes the outcome from the sensing that changes the brochure.
Why absorption is a dose problem
Serum delivery is a dose problem in the same way phototherapy is: the result depends on how much active ingredient reached the skin, over what area, for how long. A device that sprays or infuses without measuring is relying entirely on the user to judge the dose.
Users are bad at this. They over-apply when the product feels pleasant and under-apply as novelty fades. Both reduce the chance that the product does what it claims.
The three sensing functions that matter
Contact detection. The device knows the head is on skin. This gates output, which matters for safety and for dose consistency. It is cheap, robust and already standard in good devices.
Dose tracking. The device counts delivered volume or treatment time and stops at a limit. This is firmware, not sensing, and it is the single highest-value “smart” feature in the category.
Adaptive output. The device changes flow or intensity based on a measurement — impedance, temperature, or reflectance. This is where claims currently outrun evidence, because the correlation between the measurement and the ideal dose is not established.
| Function | Cost | Evidence needed | Verdict |
|---|---|---|---|
| Contact detection | Low | None beyond validation | Ship it |
| Dose tracking | Low | None | Ship it |
| Adaptive output | Medium | Real correlation data | Only with evidence |
Quantitative discharge: the mechanism that actually controls dose
Dose control does not have to start with a sensor. The most reliable version we build is mechanical: a spring-loaded pump that discharges a fixed volume per actuation.
It changes the product in three ways:
- The dose is repeatable regardless of how hard the user presses
- The count is known, so the firmware can cap a session without estimating
- The cartridge empties predictably, which makes the last-fill behaviour consistent
The trade-off is that the user has less control over flow, and the mechanism adds cost and a wear part. For an infusion device where consistent delivery is the whole promise, that is usually the right trade. For a device where the user wants to vary application, it is not.
The practical rule: decide whether the product is selling control or consistency, because the mechanism follows from that and not the other way round.
What sensing does not fix
Two things are commonly assumed to follow from sensing, and neither does.
It does not fix a bad fluid path. If the delivery mechanism is inconsistent, measuring the inconsistency does not remove it. Sensing tells you the dose is wrong; the engineering fix is in the path.
It does not create personalisation. A device that adjusts to a measured skin property is still applying a fixed algorithm. Calling that personalised treatment is a marketing decision, not a technical one — and in several markets it is a claim that regulators now look at.
The certification consequence
Adding sensing and control widens the EMC scope and adds firmware validation work. A device that adjusts output based on measurement has to fail safely, and demonstrating that is a documented activity, not a claim in a brochure.
Practically, this means the smart version of a device costs more to certify than the dumb version, and the difference is larger than most product plans assume.
What to specify
- Which sensor, and what it measures
- What the firmware does with the reading, including the failure default
- Whether dose is tracked per session and whether it is stored
- What the manual tells the user about the sensing
- Which certification scope the added electronics requires
Where it is going
The realistic next step is not more sensors — it is a device that reliably delivers a known dose, records what it did, and lets the user see it. That is achievable now, and it makes outcomes repeatable.
We build infusion and facial platforms with contact detection and dose control, and we say plainly which models carry which. The microneedling and infusion range lists the current platforms, and technology covers the underlying mechanism.
Data, logging and what it obliges you to
Once a device logs usage, it is handling data, and that changes the product’s obligations in ways that product teams routinely miss.
On-device logging only. The device counts sessions and doses and stores them locally. No personal data leaves the device, no privacy notice is needed beyond the manual, and the feature is genuinely useful for the user.
Phone connectivity. As soon as usage data goes to an app, the data is personal in most jurisdictions, and the obligations start: a privacy notice, a lawful basis for processing, retention rules, and a deletion path. In the EU this is GDPR territory; the app store reviews add their own requirements.
Cloud sync. Adds a processor relationship, security obligations and a much longer compliance tail.
For most device projects the first option delivers most of the user benefit at a fraction of the compliance cost. The second is worth doing when the app itself is part of the value — coaching, protocol progression, community — rather than simply a record of usage.
We would rather say this at the specification stage than after the app has been built, because the architecture decision is hard to reverse.
Specifying a connected device? Send us the sensing requirement and the target markets. We will tell you what the sensing can support and what it will add to certification.