The context
How do you get a child from Medellín interested in the bat, the motmot, the opossum or the glass frog — the animals they share a city with without knowing it? That was the challenge of this academic project done by a team of 3: designing “Following the Blue Trail”, a temporary interactive installation proposed for Parque de la Conservación, aimed at children aged 8 to 11 and connected to the concept of Effective Public Space. I led the team and took on the complete working prototype, the render of the space and much of the project’s development.
The full proposal defines a space of 8.5 × 9.75 meters where blue tracks on the floor guide the child to a monitor with physical buttons for choosing an animal; the choice transforms the entire atmosphere — light, sound — and activates minigames on touchscreens that let the child live what that animal experiences in the city. The journey is designed as an emotional arc in five stages: from the curiosity of the entrance to the inspiration of the exit.
The key design decision
An installation like this costs tens of millions of pesos. Before proposing that investment, we had to answer the uncomfortable question: do the interactions work? The decision was to validate cheaply what would be expensive to build: a working prototype with an ESP32 simulating the physical button panel and a minigame controlled by hand gestures, using computer vision with OpenCV on top of Streamlit. With that we could put real people in front of the space’s two central interactions without building the space.
The validation
We ran tree testing across two sessions with 10 participants at EAFIT’s Medialab, with a structured script, informed consent and video recording. A limitation we reported transparently: because of scheduling logistics, the participants were university students and not children — enough to evaluate the information architecture, not enough to consider the children’s experience validated.
The results gave a revealing contrast:
- Physical buttons: 80% success. Navigating the animals’ information flowed, with minor adjustments (icons on the buttons, relocating the back button).
- Gestures: 50% success. Half the participants couldn’t control the game. The gestures weren’t intuitive and permanent visual cues were missing — two issues of serious severity that we documented along with their recommended redesigns.
The outcome
The project closed with a proposal ready to be implemented: an executive summary, an assembly sheet with plans, an equipment list, a two-week schedule and an estimated budget — and with a prototype that had already shown what worked and what didn’t. The most “innovative” interaction (the gestures) turned out to be the most fragile, and the simplest one (physical buttons) the most solid: exactly the kind of finding that justifies prototyping before building.
What I learned
This project left me with a bittersweet lesson: we never managed to test with children, our real users. Validating with university students gave us valuable findings about the architecture and the interactions, but the most important question — does a 10-year-old enjoy this? — stayed open. I learned that access to the right users is arranged from day one of the project, with the same priority as the prototype; if you leave it for the end, you end up validating with whoever you can and not with whoever you should.
And the data confirmed something I now apply to everything I design: the simplest interaction beat the most novel one hands down. Innovation isn’t in the sophisticated gesture, it’s in it working.
