Home / UI/UX / To the point!

How do you turn exhibition visitors into outspoken contributors?

And how would such an experience look, feel, and function within the context of a historical exhibition about radio in Austria?

Tune in and find out
VISIT ONLINE

To the point!

Listen and join the conversation

An interactive application that lets exhibition visitors record audio clips as response to burning questions. Transcribing and translating speech in the background, AI helps with overcoming accessibility, language, and editorial difficulties.

Roles: Technical concept, user interface & experience design, project management

Look and feel

Digital experiences in history museums often look a bit antiquated themselves. But must they? Here’s an expressive design that’s fun to use, intriguing to explore, and—most of all—inviting to contribute.

  • Exhibition design

    The application feels as a virtual extension to the physical exhibition: Shapes, colors, and icons were highly informed by the overall aesthetic.

  • Visual metaphors

    Derived from the application’s title, the point (circle, dot) is representative of the visitor as a sender of signals. Airwaves, frequencies, tuners etc. become visual metaphors in the user’s journey.

  • Touch first

    Besides speech, the main interaction by a user with the interface is touch. Buttons are sized appropriately and react responsively to signal system feedback.

  • Playful details

    Loving an item, for example, adds a little heart-icon to its button in the navigation bar.

DESIGN CHALLENGE #1

Making speaking in public less awkward

Let’s face it: Even in times of AI-powered assistants, talking to a computer can make people feel self-conscious. All the more so in public where others can hear them.

Asking museum visitors for their “voice,” then needs to take place in an environment suitable for such a delicate request. For a couple of reasons:

  1. privacy (create space for sharing opinions without eavesdropping)

  2. sound quality (reduce background noise)

  3. consideration (limit distraction for other visitors)

Solution: Place the action in a recording booth that meets these requirements while attracting attention.

DESIGN CHALLENGE #2

Overcoming language barriers and streamlining editorial work with AI

During early planning, the project team raised an intriguing question:

What if visitors could use their own language rather than being required to choose either German or English—languages they do not know or do not feel comfortable speaking?

It turns out visitors will respond in their favorite language, creating a diverse soundscape of opinions! With translations and transcriptions generated by AI, the application manages to highlight linguistic diversity while still improving on understanding and accessibility. Editorial work is thus limited to observing content guidelines and basic proof reading.

Solution: Use AI for translations and transcriptions to decrease editorial work and increase understanding and accessibility.

The application automatically detects the language spoken and invokes AI translation and transcription in the background; these processes are not visible to the user though. (And yes, I do speak enough Polish to trick the system.)

An audio message originally recorded in Polish with a German translation generated by AI (listen to Piotr’s audio).

An audio message originally recorded in Chinese and transcribed by AI; note the disclaimer at the bottom (listen to Julia’s audio).

Advanced prototyping and user-testing

The software Figma enabled user-testing with both low- and high-fidelity prototypes in every stage of the design process: from testing basic navigational structures to the vastly more complex interaction of recording speech.

Early proof of concept

Some screens didn’t make it into the final version, others were redesigned, but the basic user-journey was already in place.

The very first low-fidelity prototype mapped the basic user journey. Even though the design changed substantially until launch, early user-tests had shown basic navigational structures were fully understood.

Recording speech

1

2

Starting the recording immediately after pressing the button seemed to be too rapid. As a result, users wasted precious time realizing recording had already started (note the gap at the beginning of the waveform in step 3).

3

Recording speech (revised)

1

NEW

Adding a simple countdown transition after pressing the recording button solved the issue: It clearly signalled to the user the recording was about to start.

2

3

Conclusion

Designing the application and seeing it come to life during development was an amazing experience. The project team rejoiced when the first genuine visitor contributions in languages other than German or English were submitted.

During the exhibition’s run between January 31, 2025 and January 11, 2026 there were:

5

questions asked

19

languages spoken

309

voices recorded

To the point!

Listen and join the conversation

Visit online