When Clothing Starts to Listen: What Wearable Biosensors Can—and Can’t—Tell Us

Estimated reading time: 7 minutes.

You finish a walk, glance at your wrist, and see a small collection of numbers: steps, heart rate, pace, perhaps a recovery score.

It can feel as though the device has understood the experience.

But every number began as something simpler: light reflected through skin, movement captured by an accelerometer, electrical activity, pressure, temperature, or another signal that software translated into an estimate.

Now imagine those sensors moving from a watch into a shirt, sock, sleeve, or patch. That is the promise of wearable biosensors and electronic textiles: clothing that does more than cover the body. It gathers information while the body moves.

The field is advancing quickly. It is also easy to describe with more certainty than the evidence allows.

Here is a practical way to understand what smart clothing can measure, where the difficult work begins, and what to look for before trusting the result.

Start with the signal, not the claim.

A wearable sensor does not begin with a conclusion such as "you are fatigued" or "you need more recovery." It begins with a physical or chemical signal.

Depending on its design and placement, a wearable system may detect:

  • Movement and pressure, including steps, joint motion, foot contact, posture, or changes in fabric strain.
  • Electrical activity from the heart or muscles, often through electrodes that need stable contact with the skin.
  • Optical signals, such as the light-based measurements commonly used to estimate pulse.
  • Temperature at or near the skin.
  • Chemical information from sweat or other body fluids, including pH, electrolytes, or metabolites in experimental systems.

These are not interchangeable. A pressure sensor woven into a sock and an electrochemical sweat sensor may both be called wearable technology, but they measure different things and face different sources of error.

A 2026 review in npj Flexible Electronics describes major progress in flexible materials, multimodal sensing, and real-time analysis. It also identifies persistent challenges: motion artifacts, environmental interference, power supply, long-term reliability, privacy, and clinical interpretation.

That last distinction matters. Detecting a signal is one task. Turning it into a dependable conclusion is another.

Fit is part of the instrument.

With ordinary clothing, fit influences comfort, movement, and appearance. In sensor-equipped clothing, it can also influence the reading.

An electrode may lose contact as a shirt shifts. A strain sensor may sit differently on two bodies. Sweat can improve contact in one part of a system while disrupting another. Repeated motion can create noise that resembles the signal researchers are trying to measure.

This is why the garment itself cannot be treated as a neutral container for electronics. Pattern shape, compression, seam placement, stretch, and sensor location all become part of the measurement system.

Even an impressive prototype may work only under specific conditions: on a particular body area, during a defined activity, for a limited time, or with careful calibration. Those limits do not make the research unimportant. They tell us what has actually been demonstrated.

One recent example shows both the potential and the caution required. A 2026 Nature Communications study described an electronic-textile patch designed to manage changing sweat levels while combining chemical and electrophysiological signals. Its model used more than 200,000 time-based observations, but those observations came from only 10 participants. That is promising prototype work, not proof that the same system is ready for every body type, activity, or retail setting.

A number is not automatically an answer.

Once a sensor records a signal, software usually filters, combines, and interprets it. This is where useful patterns can emerge and where assumptions can become difficult to see.

Consider fatigue. It does not have one universal wearable signal. Heart activity, muscle signals, motion, temperature, sweat chemistry, sleep, effort, environment, and personal history may all contribute to an estimate. A system trained on a small or narrow group may not behave the same way for people outside that group.

For readers, the useful question is not simply, "Does it use AI?" It is:

  • What does the sensor directly measure?
  • What does the software estimate from that measurement?
  • Who was included in the validation?
  • Was the device tested during real movement or only under controlled conditions?
  • How large was the error, and what happened when the device lost a clean signal?

Clear answers help separate a useful tool from a confident-looking dashboard.

Washability has to mean more than "survived a wash"

Clothing bends, stretches, absorbs sweat, rubs against skin, and goes through laundry. A sensor that works once on a laboratory sample still has a long journey before it behaves like dependable apparel.

Researchers are making real progress. In a 2025 Nature Communications study, printed sensors on cotton fabric were tested through bending and up to 10 standardized wash cycles. Performance remained comparatively stable through six washes, then declined further; by the tenth wash, the measured response had fallen by as much as 38% from its pre-wash level.

Another 2025 study of conductive knitted fabrics found that knit structure affected electrical performance and durability. After repeated artificial perspiration and washing cycles, the samples showed oxidation, delamination, changes in resistance, and, in one construction, major dimensional instability.

These findings are useful because they show the trade-offs, not because they settle the question. Ten wash cycles are not a garment's full life. A strong result in one textile construction does not automatically transfer to another. Real-world care also varies by detergent, temperature, drying method, abrasion, and how often a garment is worn.

When evaluating a smart garment, look for the test method, number of cycles, care instructions, and performance after testing, not simply the word "washable."

Wellness and medical use are different promises.

A wearable can be useful without being a medical device. It may help someone notice trends, record activity, or understand how their routine changes over time.

That does not make every reading suitable for diagnosis or treatment.

The boundary becomes especially important when a product claims to measure something that could influence medication or urgent health decisions. In a February 21, 2024 safety communication, the U.S. Food and Drug Administration warned against smartwatches and smart rings claiming to measure blood glucose without piercing the skin. The agency said it had not authorized, cleared, or approved such devices to measure or estimate blood glucose on their own.

The broader lesson is simple: the seriousness of the claim should determine the strength of evidence you expect. A general activity estimate and a measurement used to guide medical care do not require the same level of trust.

Ask what happens to the data.

Smart clothing may feel more passive than a phone, but the information it collects can be highly personal.

The Office of the Privacy Commissioner of Canada notes that wearables can collect information about a person's condition, activities, and everyday choices. Its guidance recommends checking what data is collected, how it is used or shared, how long it is retained, and whether the device offers meaningful privacy settings.

Before connecting any wearable, consider:

  • Can it work without continuous cloud storage?
  • Can you delete or export your data?
  • Does the company share information with advertisers, insurers, employers, or other third parties?
  • Will the product still function if its app or subscription disappears?
  • Can you remove the sensor, battery, or electronic module for washing, repair, or end-of-life handling?

Privacy, repairability, and product life are not side issues. They are part of whether the technology belongs in clothing at all.

Read smart clothing with a clear eye.

The most interesting future for wearable technology may not be clothing that produces more numbers. It may be clothing that measures fewer things well, explains uncertainty clearly, protects the person wearing it, and remains comfortable and useful over time.

At Arcadia, we believe human-centred technology should earn trust through evidence, thoughtful design, and honest limits. A prototype can point toward a remarkable future without pretending that future has already arrived.

So, when a garment says it can listen to the body, begin with three questions: What is it sensing? How was it validated? And who controls the information?

Those answers tell you far more than the word "smart."

Move forward.

Sources & further reading

Sources reviewed October 2, 2026. The cited research includes reviews and experimental prototypes; it does not establish the performance of Arcadia products or confirm that any described capability is commercially available through Arcadia Apparel.

0 comments

Leave a comment