Proj S01 – Smart Home Device For Quality Milk Beverage

Rethinking Milk Frothing for the Modern Home

Milk‑based coffee beverages rely on high‑quality microfoam, yet home users often must choose between manual steam wands with a steep learning curve and convenient automated frothers that compromise quality. This project reimagines the home milk frother as an intelligent, adaptive device that enables users to consistently produce cafe‑quality microfoam. By combining sensor data with AI‑driven milk quality evaluation, the system supports different milk types while minimising milk wastage.

Team members

Lim Yee Kiat (EPD), Xiong Qian Long Mike (ISTD), Lim Wei Xuan Jonas (ISTD), Lim Sherri (ESD), Long Yan Ting (ESD), Reese Ng (EPD), Stephanie Yong Jing Yi (EPD)

Instructors:

  • Zhao Fang

Writing Instructors:

  • Susan Wong

Great Coffee isn't just about the Beans

Cafes are the gold standard for a reason.

Beyond quality beans and precise brewing, baristas shape the texture of milk by carefully introducing air during frothing.

This transforms milk into a smooth, velvety finish known as microfoam.

Only when the bubbles are micro enough, will the foam be silky and shiny.

This will give the drink a rich and creamy mouthfeel.

– Jervis Tan, Two-time World Latte Art Champion

Microfoam is what brings your drink to the next level by giving it a rich and creamy mouthfeel.

When done right, it’s not just milk – it’s an experience

– Ashley Chan, Head of Coffee at Apartment Coffee, Singapore

Images from Instagram by @baristartist (2024) and @apartmentcoffee (2022)

Why Home Frothers Don’t Measure Up

While coffee brewing equipment has evolved rapidly, innovation in milk frothing has lagged behind.

We spoke to real users (baristas, home cafe owners, and coffee drinkers) about their struggles.

Hassle of Cleaning

Improper cleaning of components results in the accumulation of milk residue. This leads to hygiene issues, odour, and degraded performance. The hassle of cleaning discourages ownership and regular use of frothers at home. 

Poor Compatability with Alternative Milk

Milk from different brands and milk types have varying fat and protein compositions, affecting its aeration behaviour. While some frothers offer presets for alternative milk, they often lack the adaptability required for consistent results.

Steep Learning Curve

Frothing outcomes vary widely with many factors, such as technique, temperature, and even the choice of pitcher. Users across all skill levels face challenges in reproducing consistent, cafe-quality microfoam.

High Milk Wastage

Milk wastage commonly occurs during practice, whether from failed frothing attempts, or from preparing more milk than needed. For home users, this waste translates directly into cost, reducing their willingness to practice.

AutoMilk

Enjoy Better Quicker Smarter Milk Frothing at Home

AutoMilk is an automatic milk frother that delivers cafe-quality microfoam at home with minimal effort.

It simplifies the frothing process by using real-time sensing and AI-driven evaluation to ensure consistent results across all milk types.

Pour milk into the milk tank

Wait for milk to be frothed

Try again if calibration required

Enjoy cafe-quality microfoam!

Designed for Effortless Microfoam

AI Foam Evaluator

At the core is a camera-based AI model, supported by an adaptive light source that ensures robustness under any lighting conditions.

Embedded Sensors

A Time-of-Flight sensor evaluates milk expansion during frothing, estimating the level of aeration.

Adaptive Frothing Mechanism

The automatic frothing mechanism adjusts dynamically based on signals from the AI model and sensors, continuously improving its output.

Trusted By Baristas. Loved at Home

AutoMilk was validated with 25 baristas and skilled coffee enthusiasts.

Here’s what they had to say.

Would use AutoMilk again

Very Easy!

Average Rating for Ease of Use

★ ★ ★ ★ 

Average Rating for Foam Quality

Project Acknowledgements

Morning Logo

We would like to express our sincere appreciation to our industry partner, Akronym Pte Ltd (Morning), for entrusting us with this project and for their continued collaboration and support. In particular, we extend our gratitude to Mr Bowen Chiou, Mr Leon Foo, Mr Sullivan Su, and Mr Kevin Lim for their mentorship and insights.

We are also deeply grateful to our academic mentor, Dr Zhao Fang (Senior Lecturer, Engineering Systems and Design, SUTD), whose expert guidance, supervision, and advice were instrumental in shaping our research methodology approach.

Finally, we would like to thank Dr Susan Wong (Centre for Writing and Rhetoric, SUTD) for her constant support and constructive feedback on our communication and presentation strategies, which provided valuable fresh perspectives.

© 2026 Akronym Pte Ltd

Menu

ornament-menu