I am exploring haptic responses that could support regulation without requiring the traveller to look at a screen or interpret a complex notification. This paper interested me because the device does not simply buzz to attract attention. It delivers a slow, heartbeat-like tactile rhythm intended to influence arousal more passively.

This gives me another useful example of a wearable where the response itself is bodily and rhythmic.

It is also useful for my current biosensing questions because the rhythm is personalised using the wearer’s resting heart rate, but the system tested here is not continuously adapting to live physiology.

What doppel is

doppel is a wrist-worn wearable that produces a double, heartbeat-like tactile pulse against the inside of the wrist.

The device was designed to be discreet, on-demand and user-controlled.

In the commercial concept described in the paper, the wristband connects to a smartphone app. The user measures their resting heart rate and can choose the speed and intensity of the tactile rhythm in relation to that baseline.

The idea is based on entrainment: the possibility that external rhythms can influence bodily or emotional states.

For the calming condition tested in this study, the tactile heartbeat was deliberately set slower than the participant’s own resting heart rate.

How the rhythm was personalised

Each participant completed a resting baseline while heart rate was measured.

The active device was then set to approximately 20% slower than that participant’s resting heart rate, with upper and lower limits imposed so that the rhythm did not feel unnaturally fast or slow.

The average tactile rhythm was around 58 beats per minute.

What I find interesting here is that the output is personalised using biodata, but only at the beginning.

The system does not keep reading the user’s heart rate and dynamically chase it. Instead, physiology is used to create an individual starting point for the tactile response.

That is a much simpler form of bioresponsive design than a continuously adaptive closed loop.

What the researchers tested

The study involved 52 adults, randomly assigned to two groups.

Twenty-five participants were in the active doppel condition and 27 were in the control condition.

Both groups wore the device and were told that it was measuring blood pressure. Participants were not told that the actual purpose was to test the effect of tactile rhythm.

They were then asked to prepare a five-minute speech that they believed they would shortly deliver to an unfamiliar audience.

This was used as an anticipatory social-stress task.

The active group received the slow heartbeat-like vibration during the preparation period. The control group wore the same device but it was switched off after an initial brief vibration.

The researchers measured skin conductance, heart rate and self-reported state anxiety.

What the results showed

The public-speaking task successfully increased arousal and anxiety across participants.

However, the active doppel group showed a smaller increase in skin conductance than the control group.

They also reported lower state anxiety after speech preparation and rated the task as less stressful.

This gives the paper evidence that the active tactile condition was associated with reduced physiological arousal and lower reported anxiety during this particular anticipatory-stress task.

However, heart rate itself did not differ significantly between the groups.

This is important because the paper should not be simplified into “the slower heartbeat slowed the wearer’s heart”. That was not what the measured heart-rate result showed.

Why the passive quality interests me

Unlike the breathing-guidance paper I looked at previously, participants were not instructed to synchronise their own behaviour with the vibration.

They were not asked to breathe at a particular pace, count pulses or consciously regulate themselves.

The tactile rhythm could therefore operate more in the background.

That feels potentially useful for my airport project because a traveller may already be using significant attention for navigation, communication, luggage, security or unexpected changes.

A response that does not demand continuous interpretation could be more appropriate than a cue that asks them to perform another task.

However, that remains a design possibility rather than something this paper tested in an airport.

This is not live biofeedback

A really important distinction is that the experiment did not use real-time biofeedback.

The tactile rhythm was calculated from resting heart rate, but once the task began the vibration was not continuously controlled by the person’s ongoing cardiac signal.

The authors explicitly distinguish their method from biofeedback systems where external stimulation reflects live physiological activity.

So, in this study:

resting heart rate → sets slower tactile rhythm → rhythm remains fixed during the intervention

rather than:

live heart rate → continuously changes the tactile rhythm

For my project, this is useful because it shows that meaningful personalisation may not require constant sensing.

What could biodata control?

This gives me another possible answer to my question:

What could biodata actually control?

Here, biodata controls the baseline tempo of the tactile response.

A similar system could potentially use physiological information to set a personalised rhythm before use, while the traveller controls whether they activate it.

A more advanced design could theoretically change timing, intensity or duration in response to live data, but this paper does not demonstrate that.

It therefore gives me a strong simpler comparator for the fully bioresponsive idea.

If a resting baseline and manual activation can already create a personalised response, then live sensing would need to justify what additional value it provides.

A major limitation in the comparison

The control condition is important.

The study compared slow vibration with no ongoing vibration.

It did not compare the slow heartbeat rhythm with a faster rhythm, a neutral rhythm or another form of vibration.

This means the study cannot fully establish whether the benefit came specifically from the slow heartbeat-like tempo, from rhythmic vibration more generally, or from simply having a tactile stimulus present.

That is a significant limitation for my design interpretation.

I can say the active slow-rhythm condition performed differently from the inactive device. I should not claim that the study proves the slower-than-heart-rate rhythm was uniquely responsible for the effect.

Commercial involvement also matters

The study was commissioned and funded by Team Turquoise Ltd, the company developing doppel.

Several of the authors co-owned the company, and another author was an independent consultant for it.

That does not invalidate the study, but it is relevant when I assess the strength of the evidence.

For my project, I should treat this as a promising experimental precedent rather than relying on it alone to justify heartbeat-like stimulation as an established intervention.

Autism and sensory relevance

This study did not involve autistic participants.

It therefore cannot tell me whether heartbeat-like wrist vibration would be comfortable, calming or regulating for autistic adults.

That matters especially because my wider research keeps showing sensory variability.

A slow repetitive rhythm might feel grounding to one person and irritating, distracting or intrusive to another.

The wearable would therefore need preference controls over whether vibration is used at all, intensity, rhythm, duration and immediate stopping.

Airport relevance

There are several reasons the mechanism could be interesting in an airport.

The wrist form is discreet and portable. It does not cover hearing or vision. It could potentially operate while walking or waiting, and the person does not have to stop and complete a guided exercise.

That makes it more compatible with orientation than interventions that demand visual attention or remove environmental sound.

But the airport context also introduces limitations not present in this experiment.

The participants were seated while anticipating a speech. Airport movement, rushing, carrying luggage and cognitive demand could affect physiology and the acceptability of the sensation.

The study also tells me nothing about what happens if the vibration activates unexpectedly in an already overloaded moment.

So the airport relevance remains a design inference, not an evidence-based transfer.

What this means for my project

This paper strengthens my interest in low-attention rhythmic haptics.

Compared with breathing guidance, doppel is interesting because the user does not have to consciously match their body to the rhythm. The tactile response can potentially sit more quietly in the background.

It also changes how I think about biosensing.

A personalised system does not automatically need to sense continuously. A simpler architecture could be:

resting baseline → preferred tactile rhythm → manual or prompted activation

rather than:

continuous sensing → continuous automated adaptation

That feels especially important because one of my open questions is whether biosensing is genuinely necessary.

This paper gives me a useful design alternative to test against a fully closed-loop system.

At the same time, the study does not prove that slow heartbeat-like vibration is uniquely calming, and it does not provide autism- or airport-specific evidence.

I am therefore keeping doppel as a strong precedent for personalised, low-attention tactile rhythm, while treating both its mechanism and its transfer into autistic airport accessibility as areas that still need testing.

Source

Azevedo, R. T., Bennett, N., Bilicki, A., Hooper, J., Markopoulou, F. and Tsakiris, M. (2017). The calming effect of a new wearable device during the anticipation of public speech. Scientific Reports, 7, 2285.

Key reading from my analysis: device concept and entrainment rationale, pp. 1–2; participant sample and personalised rhythm, pp. 3–4; skin-conductance, heart-rate and anxiety findings, pp. 4–5; discussion, interpretation and study context, pp. 5–6; funding and competing interests, p. 7.