I am interested in regulation technologies that can work without taking over the user's visual or auditory attention, especially in demanding situations where the person is already concentrating on something else.

This paper is useful because it tests a tactile breathing guide during driving, where attention has to remain on the surrounding environment. That feels relevant to my airport context. A traveller may already be navigating, watching signs, carrying luggage, listening for announcements and making decisions, so any support system needs to avoid becoming another task.

What makes this paper especially useful is that the intervention worked physiologically for most participants, but its subjective benefit varied considerably between people. That makes it less of a simple “haptic breathing works” paper and more a study about the relationship between regulation, attention, personalisation and context.

What the technology is

The researchers created a vibrotactile wristband that guides slow breathing through vibration.

The wristband contained three small vibration motors positioned against the inside of the dominant wrist. During each guided breath, the vibration gradually increased in strength. Participants were instructed to inhale while the vibration was present and exhale during the silent pause that followed.

The important part is that the rhythm was personalised rather than fixed.

Before the task, the researchers measured each participant's resting respiratory rate. The wristband then guided them towards a breathing rate approximately 30% slower than their own resting rate.

Across the sample, this resulted in guided rates ranging from approximately 5 to 15 breaths per minute.

This gives me a useful example of personalisation that does not require a complex AI model. A simple physiological baseline can already change the response the wearable delivers.

Why tactile breathing guidance is interesting

Slow breathing is usually taught through visual instructions, spoken guidance or deliberate counting.

In a demanding environment, those methods can compete with the task the person is already doing.

The researchers therefore used touch because vibration could potentially remain private, discreet and less intrusive than visual or auditory guidance.

I find that particularly relevant to airport accessibility. A tactile cue could potentially support breathing without covering announcements, adding another screen or requiring the traveller to stop what they are doing.

However, the paper also shows that “unobtrusive” does not mean “effortless”.

Participants still had to notice the vibrations, understand the inhale/exhale pattern and consciously synchronise their breathing with it.

What the researchers tested

The Methods section reports 88 adults were recruited and five were excluded because of simulator sickness, leaving 83 participants in the final sample.

Participants used a realistic driving simulator.

The experiment included automated driving, a stressful or non-stressful memory task, a takeover request and then a period of manual driving. During some manual-driving periods the tactile respiratory guidance was active, while in others it was not.

The researchers collected physiological, driving and subjective data.

They were interested in two things: whether participants could actually slow their breathing using the wristband without disrupting driving, and why the intervention appeared more subjectively helpful for some people than others.

What actually happened to breathing

The physiological effect on respiration was clear.

Without the breathing guide, participants averaged around 18 breaths per minute during manual driving. With the wristband active, this dropped to around 11 breaths per minute.

This was an average reduction of approximately 35%, closely matching the intended personalised target.

The respiratory guidance was also associated with an increase in RMSSD, an HRV measure influenced by slow breathing.

So the wristband did successfully change the physiological behaviour it was designed to influence.

That is useful evidence for the mechanism: people can follow a tactile breathing rhythm while carrying out another demanding task.

But slower breathing did not automatically mean feeling less stressed

This is the part of the paper I find most important.

Although participants successfully slowed their breathing, the researchers found no overall reduction in negative feelings across the whole group when the wristband was active.

When they looked at individual patterns, the participants divided quite clearly.

Thirty participants reported reduced negative feelings with the guidance, while 34 reported increased negative feelings. Others showed no change or had incomplete data.

So people could achieve the same basic physiological task — slowing their breathing — while experiencing the intervention very differently.

That is a strong reminder for my project that physiological efficacy and subjective usefulness are not the same thing.

The attentional cost really matters

Several participants said that following the vibration required concentration.

Some found it easy, while others described the wristband as an additional demand alongside driving. When the driving situation became more complex, some participants prioritised the road and found it harder to follow the breathing guide.

The quantitative results reflected this as well. Greater difficulty synchronising with the wristband was associated with a lower probability of being in the group whose negative feelings decreased.

This gives me a very useful design principle:

A regulation cue can become counterproductive if the user has to spend too much attention interpreting or following it.

For an airport traveller, this could be even more important. Support may be needed at exactly the point when attentional resources are already stretched.

Personalisation needs to go beyond baseline

The system already personalised the breathing rhythm using each participant's resting respiratory rate.

Even so, some participants still reported that the vibrations felt too fast or were difficult to coordinate with.

The researchers therefore recommend giving users more control over vibration intervals and allowing slower rhythms to be selected or adapted.

This shows that physiological personalisation is not necessarily the same as experiential personalisation.

A technically appropriate rhythm calculated from someone's resting breathing rate may still feel wrong to that person.

For my project, that suggests that biodata should not replace preference settings. The traveller may need to choose what pace, strength and style of tactile cue actually feels comfortable.

What could biodata control?

The wristband in this study was not closed-loop in real time.

It used a physiological baseline measured beforehand to set the breathing rhythm, but live physiological data did not continuously change the vibration pattern while participants were using it.

The researchers do, however, discuss the possibility of embedding sensors such as EDA or photoplethysmography (PPG) into a wristband so that a system could trigger a just-in-time intervention.

For my project, there are several plausible things biodata could therefore influence:

when a breathing cue is offered; whether the pacing needs to change; how long the sequence continues; or when support should stop being offered.

But the paper makes me cautious about fully automatic activation.

If the context is already highly demanding, adding a regulation cue at that exact moment could increase attentional burden rather than reduce it.

Context might matter as much as physiology

One of the strongest findings is that the effectiveness of the guidance depended partly on how demanding or critical the driving situation felt.

Participants who appeared to allocate more attention to the driving task were less likely to report subjective benefit from the guidance.

The authors therefore recommend context-aware interventions, where the system considers what is happening around the user rather than responding only to the body.

This connects directly to my airport project.

A physiological change during a quiet waiting period and the same change during security, boarding or a sudden gate change may require completely different responses.

A wearable that only sees physiology could miss that difference.

This strengthens my current thinking that any future responsive system may need some combination of:

personal baseline + current context + user preference + user choice

rather than simply:

high arousal = activate calming mode.

User choice appears again

One participant described the experimental use of the wristband as constraining because they were required to follow it, and said it might have been easier to use in real life if they had the choice.

The authors later recommend flexible use conditions where people can engage with the intervention at their own pace and discretion.

This fits very closely with my project.

Even when a cue is technically effective, being made to follow it can change the experience.

For an airport accessibility system, I would therefore be more comfortable with sensing prompting or offering support than forcing the traveller into a breathing exercise.

Autism and airport relevance

This study did not involve autistic participants and it was not conducted in an airport.

Its participants were licensed drivers using a simulator, and the intervention was designed around driving stress and attentional demand.

I therefore cannot use it as evidence that tactile breathing guidance will regulate autistic adults during airport travel.

Its relevance is more specific.

It demonstrates that a tactile wearable can successfully guide breathing during a complex ongoing task, while also showing that individual differences, attentional demand and context can determine whether the same intervention feels helpful or stressful.

That mechanism and those limitations are highly relevant to the design problem I am exploring.

Where this could fail in an airport

The airport context may create even more competition for attention than the driving task used here.

A traveller might need to follow signage, hear instructions, monitor queues, locate documents, manage luggage or respond to unexpected changes.

A breathing cue that requires deliberate counting or constant synchronisation could become one more demand layered onto that situation.

There is also a sensory issue. Vibration is not neutral for everyone. Some people may find repeated wrist stimulation distracting, irritating or impossible to ignore.

This means the question is not simply whether the wearable can slow breathing.

The more important question may be:

can it provide useful support without becoming another thing the traveller has to manage?

What this means for my project

This paper strengthens tactile guidance as a possible response modality, but it also makes my design criteria much stricter.

The technology shows that vibration can guide a physiological behaviour while the user continues another task.

However, the study also demonstrates that successful physiological regulation does not guarantee subjective regulation.

That is a really important distinction for Grounded.

If I explored breathing guidance, I would want it to be:

optional; familiarised in advance; adjustable in rhythm and intensity; easy to ignore or stop; and sensitive to the current task demand.

I would also want to test whether continuous breathing guidance is actually necessary. The authors suggest that shorter or isolated tactile cues might reduce attentional demand compared with constant stimulation.

The paper therefore moves me away from thinking only about what response can be activated, and towards asking when should the system stay quiet?

That may be just as important in an airport as knowing when to intervene.

Source

Béquet, A. J., Jallais, C., Quick, J., Ndiaye, D. and Hidalgo-Muñoz, A. R. (2024). Road to serenity: Individual variations in the efficacy of unobtrusive respiratory guidance for driving stress regulation. Applied Ergonomics, 120, 104334.

Key reading from my analysis: tactile wristband and personalised breathing rhythm, pp. 7–9; physiological and subjective results, pp. 17–18; participant experiences and attentional demand, pp. 16–17 and 21–24; design recommendations and limitations, pp. 24–26.