Coffee Composer
Meet me at Milan Design Week 2019


Via Bergognone 34, 20144, Milan Italy

Supervised by Prof. Carola Zwick
Master Semester 1

3D printing, laser cutting, arduino, LED strips, stepper motor, NFC sticker, RFID reader

Xinyue Yang, Antonia Nandori

Coffee consumption on the go increasingly relies on self-service. Starbucks-style chain outlets make us stand in line to order; our workplace cafeteria requires us to operate an automated machine. While this optimizes the process for the organisation who sell the coffee, for the consumer, it becomes ever more tiresome. We spend longer queueing, are under pressure to speed up our order, and know very little about the product they will receive. This amounts to a dissatisfactory user experience, one which falls far short of traditional café service.

Queuing for a long time in public service

In front of a menu full of strange coffee vocabularies, user feels streesed to make a quick decision

The interaction is boring - pressing a button, and user knows little about what they will receive

Coffee Composer puts the fun back into self-service. This service design concept enriches the interaction for consumers, making the most of what humans do best: the creative and tactile parts of the process. A range of color-coded ‘tokens’ symbolize possible ingredients (such as milk and sugar), offering users greater transparency and autonomy to choose what ends up in their coffee. A glass cup makes it easy for users to comprehend the custom combination of ingredients as they are dispensed.

The selection process begins earlier, meaning queue time is used more effectively, as customers get to pick and stack their preferred elements to create their ideal hot beverage. The automated coffee machine intuitively ‘reads’ their order, converting each token into its corresponding ingredient in consecutive order. The communication between the automated machine and human user becomes more subtle and enjoyable.

Coffee Composer demonstrates that increased automation need not result in a banal and laboured interaction between user and machine - it should instead free us humans up to pursue activities we enjoy.

User Journey

User Flow

Background: Public Service / Semi Self Service

Ingredients Module System

This modular system is considered to be scalable depending on diffenrent preference, region and various coffee culture.
Almond Drink
Soya Drink
White Sugar
Brown Sugar
Hot Water
Oat Drink
and so on...

Each module represents an ingredient and has a bold-style graphic logo on it. The sideview of the modules visually makes the ratio between different ingredients much stronger. The color code serves as a supporting hint to help people recognize what module it is. They have different heights of 3mm, 5mm and 9mm.

a playfull experience of composing a drink

visually corresponded layers, it brings more transparency and knowledge to your drink

Choosing Process

The choosing process starts earlier as soon as people join the queue. With a playful choosing experience, it makes waiting time seem shorter.

She chose 2x Milk and 1x espresso, the ratio is 2:1 (milk to espresso)
It might taste similar as cappuccino

Machine Prototype

mechanism inside of the machine

Serial Communication Prototype

An embedded NFC sticker makes it recognizable

serial communication between arduino (nfc reader) and p5.js

Interacting with Machine

The machine starts to recognize the cup as soon as you take it.
The breathing LED light gives you a hint where to put your cup and modules. The biggest cup would be suggested to place onto the lowerst platform while the smallest cup will be suggested to the highest platform.
By placing the modules onto the surface, it can recognize the type and amount of the modules. In the end it maps the ratio into cup's volume.

By taking the cup - Machine recognizes the cup and gives you light signal
By placing the tokens - Machine recognizes the tokens and calculates the price

Coffee Composer Video

Exhibition at Milan Design Week 2019