Research question. How can Grounded feel warm, recognisable and individual while keeping changing airport information easy to distinguish and read?

Colour is part of the brand, but it also changes what people notice first and how they separate information. For this app, the audience is autistic adults travelling independently through environments where attention may already be shared between signs, announcements, people, luggage and the next action. I therefore need to decide not only which colours feel right to the identity, but where each colour belongs, what it might imply, and whether essential information still works when colour is hard to distinguish or the display changes.

What the research says about colour and feeling

Colour associations can inform a visual direction, but they do not predict one person’s response. Jonauskaite and Mohr’s 2025 systematic review identified 132 relevant studies published between 1895 and 2022, involving 42,266 people across 64 countries. It found recurring colour–emotion associations, but many associations were many-to-many and influenced by lightness, saturation and hue. The authors also note that most studies test associations in abstract settings; that does not show that a particular screen will make a traveller feel a particular emotion in an airport. Jonauskaite and Mohr, “Do we feel colours?”

This gives me a useful way to think about the current palette. Lilac can contribute a distinctive, imaginative and expressive identity; lime can make an action or small character feel energetic and visible. Those are design associations I want to explore, not claims that lilac is calming or lime is motivating for autistic adults. Colour meaning will also change when it is used as a tiny accent, a route line, a large screen background or text on a button. The audience and task matter as much as the hue.

The audience cannot be reduced to one sensory preference. MacLennan, O’Brien and Tavassoli’s participatory qualitative research describes autistic adults’ sensory experiences as complex across sensory domains, including vision. It supports asking about individual experience rather than prescribing a universal sensory environment. Maia’s own Critical Report similarly reports a small, self-selected questionnaire with themes around sensory environments, changing needs, independence and control. Neither source evaluates these palette options. I need to treat colour as a hypothesis to test with autistic adults, not as an autism-specific formula. MacLennan et al., “In Our Own Words: The Complex Sensory Experiences of Autistic Adults” · GDE751: Week 10: What Autistic Adults Need from Air Travel.

Accessibility is about contrast and meaning as well as hue

For the gate-change task, the most important colour decision is whether the new destination, previous destination and next action can be found without relying on subtle colour differences. WCAG 2.2’s AA contrast criterion sets a minimum contrast ratio of 4.5:1 for normal text and 3:1 for large text against its background. The calculation is based on relative luminance, so a different hue alone does not guarantee readable text. The W3C guidance also notes that thin or unusual lettering can be harder to read in practice even when a calculation passes. W3C: Contrast (Minimum)

WCAG also says that colour must not be the only visual way to communicate information, indicate an action or distinguish an element. In Grounded, “Go to gate A04” and “Previous gate B12” therefore need explicit words and hierarchy; the new gate cannot be indicated only by lime, and a red/green difference cannot be the only way to show a status. Shape, labels, position, icon and text can reinforce a colour cue. This supports a colourful identity while making the information available through more than one cue. W3C: Use of Color

Passing a contrast calculation would be a baseline, not proof that a screen is comfortable, easy to scan or accessible for the intended audience. I still need to test the actual colours in light and dark appearances, at phone scale, with enlarged text and on a real device. Apple’s guidance similarly recommends checking legibility in different appearances and accessibility settings. GDE751: Week 19: iOS and dark mode options

What I took from other visual references

The supplied character and app references show bright pink, lime, aqua, lilac, yellow, orange and blue. I am interested in their use of simple colour fields, clear silhouettes and expressive contrasts. The images are sources for visual exploration; they do not tell me what an airport user will prefer. The airport examples in my earlier brand research have a different job: they prioritise current journey information, navigation and service details. I want Grounded to retain a recognisable personality without letting brand colour compete with the operational fact. GDE751: Week 18: Brand research

The colour references helped me identify a lively family, but I did not want every colour to have equal weight in the interface. The next question became whether a bright colour should fill the screen, frame the journey, mark a character or carry an action.

The experiments and what changed

The boards below collect the main colour trials in one place. They show my visual iterations; they are not participant evaluations.

Earlier colour boards

First app-colour comparison: violet/lemon, teal/papaya and peach/cobalt.

Second comparison after the first set was rejected: cream/coral/blue, white/orange/lime, and ivory/violet/lemon.

Dark-field study: the surrounding interface stays dark while the gate-ticket surface changes.

Five-colour family study: hot pink, lime, aqua, lilac and yellow, without sky blue.

Off-white comparison: a light screen versus a light gate area on a dark screen.

Orange, yellow and green tests: each hue takes a different lead role.

1. Testing the full reference family. The first direction used hot pink, lime, aqua, lilac and yellow, after sky blue was removed from the selected character references. This kept the playful energy of the collage, but five equally strong colours created too many possible roles for one alert. I wanted the traveller to identify the gate and next action before reading the character or decorative route. Removing sky blue was a choice about making this identity more focused; it is not a claim that blue is less accessible or unsuitable.

2. Testing a colour as the whole background. I compared aqua, lilac and pink as the dominant screen field. The trials made the same layout feel very different: aqua and lilac gave the character and route a strong identity, while the pink field felt more emphatic. Maia’s earlier review said that a fully bright background was too much, so I moved away from using a saturated colour across the whole alert. This is a design response to Maia’s critique of these mockups, not a finding about autistic users generally.

3. Testing colour roles rather than adding more hues. I then compared different combinations and placements, including orange/yellow/green, off-white surfaces, a dark neutral field, and light/dark versions of the lilac-and-lime pairing. The orange/yellow/green study was vivid, but it made the background, route, gate frame, character and action all carry strong colour at once. My concern was competition and the number of simultaneous signals. I kept the individual hues as useful reference material but did not continue that combination as the main brand direction.

The light/dark comparison changed the question from “Which background colour is best?” to “Can the same identity work with different surface colours?” It keeps lilac in the route role and lime in the character/action role, while the surrounding surface and text adapt. Maia identified this as a possible dual-appearance direction. It remains provisional: the images are static studies, not tests on an iPhone or with participants.

4. Setting aside colours that did not fit the role. Pink and aqua remain part of the supplied references and can appear in character details or future illustrations, but I did not select them as the main alert background. Pink’s strong visual presence made it harder to keep the information surface quiet in that trial. Aqua remained visually appealing but competed with the lilac route when both were used as large surfaces. Yellow is useful in character references and small accents, but I do not want the gate, route and primary action all relying on similarly bright surfaces. The point is not that these colours are “bad”; it is that I need a small number of stable roles in this particular interface.

Earlier navy-and-peach experiments are development history rather than the current identity. The later mascot and iOS studies consistently return to a lime companion with lilac route colour and adaptable neutral surfaces. The character page shows how the lime body was carried forward as an identity choice, not a signal of an event’s status. GDE751: Week 19: Mascot: From Reference to Companion

How I will judge the palette

The colour variants should be compared using the same B12-to-A04 content, text size, layout and character placement. Changing several variables at once would make it difficult to know why one version reads differently. I need to ask autistic adult participants to complete practical tasks as well as describe their impressions:

• Which gate is current, and which one was previous?
• What would you do next?
• Does the decorative route or mascot attract attention before the gate information?
• How does the screen feel in the airport scenario, and would you choose a different appearance?
• Can you distinguish the same information in light and dark modes, with larger text and without colour cues?

I should look for individual and conflicting preferences instead of averaging them into a single “best” hue. I also need to inspect contrast pairings for text, buttons, secondary labels and unavailable/disabled states, and confirm that written labels remain clear when colour is not perceived as intended. These tasks remain to be done; the current mockups are not participant evidence.

Provisional brand palette

This is the palette direction the experiments currently support. It records the roles that Maia selected; it does not lock final colour values. Exact hex tokens should be set and checked against the intended type sizes, light/dark appearances and real-device rendering before the app specification is final.

| Colour | Brand role | Why it is here | Current status |
| --- | --- | --- | --- |
| Warm off-white | Light appearance surface | Leaves space around the information and avoids making every screen a saturated colour block. | Selected as the light neutral direction; exact value open. |
| Charcoal | Dark appearance surface and dark ticket option | Provides a dark neutral field against which the journey and character accents can remain distinct. | Selected as the dark neutral direction; exact value and component pairings open. |
| Lilac | Journey route and identity accent | Retains a distinctive connection to the selected visual references and gives the route its own consistent role. | Selected role; it must not be the only cue that information is a route. |
| Lime | Mascot and primary action accent | Connects the companion and action to the same identity while creating a strong accent against neutral surfaces. | Selected role; black text, focus, pressed and disabled states need separate checks. |

Pink, aqua and yellow remain secondary reference colours for illustration exploration, not status colours. They should not be reintroduced as competing large interface surfaces without another comparison. Current role selection is supported by Maia’s design review and the direction carried into the mascot and light/dark studies; it is not evidence of autistic user preference.

Taking forward: use warm off-white and charcoal as adaptable surfaces, lilac for the journey motif, and lime for the companion and primary action. Keep gate and status meaning in words. The roles are selected provisionally; exact values, contrast checks and participant feedback remain open.