After comparing the ideas and returning to the evidence, I currently think the strongest probable direction is:

A biosensing-informed contextual airport-support system for autistic adults travelling independently.

The important change is the role of biosensing. I am not trying to find a signal that tells the system “this person is overloaded.” The evidence does not justify that.

Instead, I am exploring physiological information as a timing / awareness signal that may indicate it is useful to check in.

Emerging system architecture

Biosensing + movement + airport context → cautious interpretation / uncertainty → transparent check-in → user choice → optional support

The contextual layer could include information such as where the traveller is in the airport, journey stage, movement, crowding, queues, noise or environmental conditions, disruption or gate changes, time pressure and nearby support locations.

The prompt should make uncertainty visible rather than pretending the system knows what the traveller feels.

“A change was detected while this area became busier. Would support help?”

The traveller remains the final interpreter.

What could happen next?

Contextual support

• quieter / less crowded route
• information about what happens next
• more or less detail
• disruption or gate-change support

Sensory support

• audio filtering
• visual filtering
• personally configured sensory adjustments

Optional regulation support

• preferred haptic or tactile cue
• breathing guidance
• audio or another personally selected option

At every stage, the traveller should still be able to pause, dismiss, override or request support manually.

A more focused sensing direction

My sensing research is also becoming more specific. The current working shortlist is:

EDA + PPG / heart rate + accelerometry

• EDA may help identify within-person physiological change, but it does not identify sensory overload.
• PPG / heart rate provides another physiological channel, but movement and context can also alter it.
• Accelerometry matters because movement/activity can help explain possible artefacts or ordinary physical change.
• HRV / PRV needs to be treated cautiously during movement.
• Skin temperature is currently a secondary contextual measure.
• EEG currently appears unnecessarily burdensome for a walking/airport context.

This is not yet a technical specification. It is simply a more evidence-led sensor territory than the earlier stage where I was considering many biosignals equally.

Current design position

The system may notice and combine information, but it should not decide what the traveller is experiencing. The traveller remains the final interpreter.

I am therefore probably moving forward with this architecture, but the exact sensors, interaction, response options and physical wearable form are still open.

What I need to investigate next

The contextual architecture gives me a much clearer system to design around, but I do not want to treat it as resolved simply because it is now more coherent.

Sensing + interpretation

• Which combination of EDA, PPG / heart rate and movement data is practical?
• How should a personal baseline work?
• How should movement, rushing or luggage be accounted for?
• What happens after false or missed triggers?

Airport context

• What airport/contextual data is realistically available?
• What can be processed locally rather than shared?
• How should crowding, route and disruption information be represented?

Interaction + response

• How should prompts be worded?
• How often can the system check in before becoming intrusive?
• Which sensory-filtering options remain safe and useful while travelling?
• What remains available manually at all times?

Physical design + testing

• Wearable form, material, placement and sensory comfort
• Privacy, consent and trust
• How to prototype the logic without claiming clinical validation
• What evidence would cause me to revise the architecture?

What I need to protect as I develop it

• physiological change is not equivalent to sensory overload;
• a prompt is not proof that support is required;
• one sensory or regulation method cannot be assumed to work for everyone;
• the wearable itself may create sensory or cognitive demand;
• and a personal wearable should not become an excuse for inaccessible airport systems.

Working position, not final outcome

Contextual airport support is now the probable core direction, but the system still needs focused research, interaction testing and physical prototyping before I can decide exactly how it should work.

This gives me a much clearer system to design around, but rather than treating the architecture as resolved, my next stage will test the evidence, interaction and physical form needed to determine exactly how it should work.