Testing a motion-sensitive installation with users
The ITP 30 Show brings together experimental art, design, technology, and media projects created by participants connected with NYU’s Interactive Telecommunications Program. A movement-responsive work can look effortless in a demonstration, yet live user sessions expose where gesture, sensing, and interpretation drift apart. Structured sessions turn those surprises into usable evidence.
Australian audiences add useful conditions to that process. Sydney and Melbourne venues can combine heavy foot traffic, variable daylight, warm weather, and crowded public programmes; visitors may also use mobility aids or prefer observing rather than performing. Testing must treat accessibility, consent, and public behaviour as design variables rather than administrative details.
Define the interaction before the session
Begin with an observable purpose, not a vague promise that the piece feels responsive. Decide which movements should trigger it, how soon the response should arrive, and what counts as a false activation. Write testable scenarios such as approaching from the left, holding still, waving once, or passing with a bag. The show’s project background can help connect the experiment to its stated concept.
Prepare a repeatable baseline before anyone arrives. Record the sensor field with the room empty, note background movement and interference, then confirm lighting, sound, network connection, and power. Use the same furniture and entry route for every round. A short dry run should establish stable thresholds without making the interaction so forgiving that ordinary use feels automatic.
Recruit participants and protect their data
Recruit a small, varied group through local design schools, community organisations, venue mailing lists, and existing participants in Australian public art programmes. Include people with different levels of technical experience and relationships to physical interfaces. Offer an accessible venue and a modest voucher so participation is not limited to those who can spend their own time and money.
Explain data collection in plain English and obtain informed consent before recording. Under the Privacy Act 1988 and Australian Privacy Principles, collect only the motion or video needed for the stated task, avoid persistent identifiers, and explain retention and withdrawal. Raw movement traces can reveal behavioural patterns, so store de-identified results and restrict access to the research team.
Offer equivalent ways to engage: seated reach, a lower activation zone, verbal commands where appropriate, or permission to watch. Public programmes in Melbourne and Sydney commonly acknowledge Country; build the same welcome into the test. If the installation gathers biometric or identifiable data, pause the test and seek specialist privacy advice.
Run controlled movement sessions
Start each session with a scripted welcome, consent check, and one-minute baseline. Ask participants to think aloud while trying three tasks: an intended gesture, an incidental movement, and a free exploration. The facilitator should answer questions neutrally and avoid coaching after the first attempt. Note hesitation, repeated gestures, changes in distance, and the point at which someone gives up.
Change conditions between rounds, not the system mid-trial. Test in daylight, at night, with event noise, and during normal footfall. A Melbourne gallery with glass walls may produce substantial reflected movement, while a Sydney site near a busy street entrance may attract constant foot traffic. Log every threshold or sensor change so a better-looking result can be traced to a known variable.
The ITP 30 Show’s cybernetics research offers a useful frame for this work: sensing, people, and response form a feedback loop. Each adjustment changes what the next participant notices. Test one controlled change at a time, preserve the earlier configuration, and compare behaviour rather than assuming that more sensitivity creates a better interaction.
Turn observations into evidence
Triangulate numbers with comments. Observation explains why a participant leaned closer, waved twice, or never moved, while quantitative measures reveal whether the installation responds reliably. Keep raw notes alongside the summary so unusual cases remain visible. Record these quantitative signals:
- Activation rate and unintended triggers by scenario
- Response latency from movement to visible feedback
- Attempts, retries, and time before abandonment
- Differences between first and repeated use
Then examine who experienced the behaviour. Compare results by interface height, mobility, familiarity, and environmental conditions without treating any group as a single type. Revisit edge cases such as children, crowded queues, long garments, and gestures that vary across cultures. A lower trigger rate is valuable only if the intended action remains clear and accessible. Capture these qualitative signals:
- The words participants use to describe the gesture
- Moments of hesitation, surprise, or loss of confidence
- Whether feedback is understood without facilitator help
- Suggestions for alternative ways to interact
Compare findings with the original hypothesis, then document each revision in a short change log. Re-test the most consequential change with a new participant rather than someone who has already learned the interaction. Preserve failed configurations because they explain the final behaviour and help another designer reproduce it.
Prepare the public iteration and invite action
A public iteration needs a technical rider, staffing plan, contingency procedure, and clear limits on interaction. The online ITP 30 Show can document the prototype, publish project descriptions and author credits, and connect exhibition events and RSVP details to an audience beyond the room. Related programme resources, such as scholarship and census information, can accompany those materials. For remote viewers, include images or clips showing the sensor, the participant’s movement, and the resulting response.
For an Australian launch, involve venue technicians early and check power, connectivity, sensor mounting, and insurance. Local design studios, independent galleries, public museums, festivals, and councils can be valuable commissioning partners, but proposals should fit real maintenance budgets and staffing conditions. Account for weather, transit disruptions, and high visitor numbers rather than assuming a quiet studio will resemble opening night.
Close with a concise test record covering the purpose, participant profile, consent approach, measurements, revisions, and unresolved risks. Publish the record on the project page, circulate it to interested collaborators, and open registration for the next public trial through the show’s event and RSVP details. Invite participants, artists, programmers, and venue teams to register and bring their evidence into the next collaborative build.