Cross Cultural
Como
A dual-surface digital experience that uncovers the global origins of Como's silk heritage — connecting museum objects to the countries, trade routes, and techniques that shaped them.
ROLE
Service Design, Group Project
Type
Mobile App, Horizontal Screen
CONTEXT
Politecnico di Milano (2024)
TOOLS
Figma, ML scan

PROBLEM 01/05
SOLO VISITOR GAP
Independent visits require a guide to make sense of the heritage — the museum depends on human mediation.
NO CULTURAL CONTEXT
Objects are displayed without their international origins — the Moroccan draw-loom, Jacquard, Chinese silk trade are invisible.
DISCONNECTED GLOBAL STORY
The museum's reach — Japan, China, Morocco, Turkey, Holland — is undiscoverable from exhibits alone.
STATIC INFORMATION LAYER
A4 sheets and instrument tags cannot convey the depth of 600 years of cross-cultural silk exchange.
INSIGHT 01 —
SOLO VISITORS
Solo visitors struggle to understand the heritage compared to guided visits, because right now it requires interaction with people who know about it.
INSIGHT 02 —
CULTURAL GAP
There is a gap in connecting the museum with different cultures and providing a historical context for the displayed objects — the global significance is invisible.
DESIGN APPROACH 02/05
Instead of a single interface, Cross Cultural Como was designed as two distinct surfaces serving two distinct physical contexts — a visitor with a phone in front of a loom, and a curious walk-in at the museum entrance. Both share one content layer: the cross-cultural history of Como's silk objects.
01
Personal SMARTPHONE
IN THE MUSEUM
Heritage App
KEY FUNCTIONS
Visual museum map — navigate by room
Heritage stories collection, unlocked by scanning objects
ML object recognition + QR code fallback
Spun — AI storytelling on a country's silk history
02
HORIZONTAL SCREEN
MUSEUM ENTRANCE
World Map Installation
KEY FUNCTIONS
Interactive world map — connect countries to Como
Cultural influences surfaced through heritage history
Spun — country-level silk storytelling, no download needed
QR code to download the app for the full experience
DESIGN 03/05
The horizontal touchscreen at the museum entrance runs without the app. It lets any visitor — with no phone, no download — immediately grasp that Como's silk is a global story, not just a local one.

Radial map, Como at centre
Positions the museum as the global nexus of silk history, not a provincial exhibit.
Concentric rings:
Italy → Europe → World
Mirrors the positioning diagram from the research phase.
Tap a country to open its silk story
The interaction requires zero instruction; curiosity drives the whole flow.

Two-option home
Binary choice removes all cognitive load for a visitor who has just walked in; no menus, no orientation needed.
ITA / ENG toggle
Foreign tourists are a core audience; language selection is surface-level, not buried in settings.
QR to app persistent
Every screen on the horizontal display shows the app download CTA; the installation converts walk-ins into app users.

WHY SPUN AS A NAMED FEATURE
Giving the AI storyteller a name and identity transforms it from a search tool into a guide. Visitors engage more readily with a named, characterised voice — especially in a cultural institution context where trust matters.
THE CONSTRAINT
The horizontal screen is a shared public installation. Spun's content had to work without personal accounts or saved sessions — every interaction starts fresh, which shaped the stateless conversational model.
The app is built around a single interaction: scan a machine, unlock its story. Every screen answers one question — where did this object come from, and why does it matter to the world beyond Como?


Lock metaphor for
locked stories
Creates a sense of discovery without suggesting the visitor is excluded.
Room-based navigation mirrors physical layout
Visitors can orient themselves without cognitive overhead.
Object recognition via a pre-trained ML model is the ideal interaction
Frictionless, no signage required. The QR label is the reliable fallback for low-light rooms, glass cases, and non-standard machine angles.

White card on dark bg
The story "surfaces" from the dark room, reinforcing the moment of discovery.
Collection tab separate from map
Visitors who want a list view can switch; map-first respects the physical context.
Spun as persistent bottom CTA
Not buried in menus; researchers can always access deeper AI-powered content.

Spun as persistent bottom CTA
Not buried in menus; researchers can always access deeper AI-powered content.
THE ALTERNATIVE REJECTED
NFC tags — rejected due to museum infrastructure constraints and the physical positioning of machines, which made reliable tap-placement impractical for all visitor heights.
DESIGN System 04/05

Neue Montreal.
8-point grid.
3 core tokens.
Neue Montreal was chosen for its confident, curatorial personality — the right balance between editorial precision and approachability for a museum audience. The 3-colour system (yellow, black, white) is strict: yellow is reserved exclusively for interactive elements and active states, keeping affordances unambiguous even on a noisy museum floor. An 8-point grid ties both screen formats together — portrait mobile and landscape 4:3 entrance screen — ensuring spatial consistency across two very different physical contexts.
Reflection 05/05
WHAT WORKED
The dual-surface architecture solved two completely different contexts without fragmenting the content — a solo phone-user inside the museum and a group of walk-ins at the entrance both access the same cultural layer. The lock / unlock metaphor created a discovery arc without requiring any onboarding. The strict 3-colour system kept the dark UI legible and the yellow affordances unambiguous in unpredictable museum lighting. Positioning Como as the nexus on the world map — rather than a destination on it — reframed the entire museum narrative.
WHAT I WOULD DO DIFFERENTLY
Run field observation sessions with solo visitors before finalising the scanning flow — the ML recognition model was designed top-down, and real-world museum lighting, glass cases, and visitor posture would have revealed edge cases earlier. I would also test Spun's content depth against visitor attention spans in situ — the AI storytelling feature is the project's most ambitious element and the one most at risk of being too long for a museum-floor context. A museum staff co-design session was missing; the content layer was reconstructed from labels and existing documentation rather than curators.