I had reached a point where generating more technology-led ideas was no longer helping me decide what the project should become. The project is still focused on supporting autistic adults travelling independently through airports, particularly around sensory demand, grounding and regulation, predictability, control and being able to move through the airport journey with appropriate support.

I therefore moved from broad ideation into a more critical comparison of the concepts I had already generated. The aim was not to choose the idea I liked most, but to understand which directions were most clearly connected to the airport problem and which questions still needed evidence.

An important design decision has also been made at this stage: biosensing is now a fixed part of the design direction. This is a change from my earlier position, where I was still questioning whether biosensing should remain at all. I am fixing it as an area of design investigation, not because new evidence proves that physiological data can identify sensory overload, anxiety or stress. I am treating physiological change as one possible input that may help a system notice change and offer timely support, while interpretation remains uncertain and the traveller remains in control.

Following critical comparison of the initial concepts, the design space was narrowed to three biosensing-informed directions: personal sensory control, contextual airport support, and user-selected grounding/regulation.

Red-teaming the original ideas

Rather than simply selecting my favourites, I used a red-team critique to challenge the ideas from the opposite direction. For each concept I asked whether the technology genuinely addressed an airport problem, whether I was overinterpreting what biosensing could tell me, and whether the intervention could introduce new sensory or cognitive demands.

I also challenged the ideas around false or unwanted triggers, privacy and trust, airport safety and situational awareness, what already exists, and whether the concept could realistically be prototyped within the MA. This shifted the question from “what could I make?” to “what role does this technology actually have, and what would justify that role?”

The critique also reinforced a principle that now sits across all three directions: a physiological signal should not automatically produce an intervention. The more appropriate interaction is closer to physiological change → interpretation/context → transparent prompt or feedback → user choice → optional response.

Three directions I am taking forward

1. Bioresponsive Personal Sensory-Control System

This direction combines physiological sensing with adjustable visual and/or auditory filtering through a wearable or head-worn system. Possible responses could include light or tint adjustment and audio attenuation.

The key point is that the system should not automatically decide that the traveller needs a “calmer” environment. Physiological change may contribute to a prompt, but the traveller should be able to accept, reject, pause or alter the response. The main questions are whether filtering can reduce sensory demand without compromising awareness, and whether the wearable itself could become another sensory burden.

2. Bioresponsive Contextual Airport-Support System

This direction combines biosensing with information about the airport journey, such as journey stage, crowding, queue conditions, disruption or route choices.

Instead of claiming to know what the traveller is feeling, the system may use physiological and contextual information together to decide whether it is useful to offer support. This might include a quieter route, a preparation prompt, simplified information or an explanation of what happens next. I am interested in this direction because it connects the responsive system directly to the airport environment rather than locating the whole problem within the person.

3. Bioresponsive Grounding / Regulation Support

This direction explores physiological change as a possible trigger for an optional grounding or regulation prompt. The traveller could select their preferred response in advance, such as haptic or tactile feedback, breathing guidance, audio or another personally useful method.

I do not want to describe any of these responses as universally calming. A response that feels supportive to one person may be distracting, uncomfortable or intrusive to another. The design question is therefore not simply which modality is “best”, but how preference, timing, intensity, duration and immediate stop/override are built into the experience.

What changed after the critique

Several earlier concepts are now parked, merged or rejected in their current form, including full VR sensory-room concepts, AI therapeutic voice support, breathing objects, chest/head vibration and thermal regulation ideas.

I do not see these as failed ideas. They were useful exploration because each exposed a different design issue. VR raised questions about isolation and situational awareness. Voice and breathing concepts raised questions about cognitive demand and existing precedents. Vibration and thermal ideas made me question whether I was assuming a particular physiological or regulatory effect without enough evidence.

Some of these mechanisms may still reappear later as components, precedents or comparison points, but they are no longer equal candidate solutions. This stage therefore represents a narrowing of the design space, not selection of a final product.

The most important change in my thinking is that I am no longer comparing unrelated technologies. I now have three coherent directions that all share the same biosensing-informed foundation but differ in what the system responds with: changing sensory input, changing contextual support, or offering user-selected regulation support.

Next steps

The next stage is targeted literature and precedent research into the three surviving directions. I need to investigate what physiological signals can reasonably contribute, especially in moving and real-world contexts, without treating them as direct measures of internal state.

I also need stronger evidence around autism-specific wearable acceptability, sensory filtering, contextual airport support, grounding/regulation modalities, and the design of uncertainty, prompting and user control.

This research should let me compare the three directions on the same basis before moving further into interaction, form and prototype development. The aim is to narrow again through evidence and testing rather than locking the project into a final product too early.