My task was to combine an emerging technology with a product.
I wanted to learn more about artificial intelligence, and decided to apply an AI based solution to a problem in the kitchen.
Through interviewing cooks and researching current household products, I narrowed in on a problem centered around food waste.
Timeline: 3 weeks, October 2018
Role: Designer & Researcher
Skills: User-testing, Prototyping, Illustrator, Photoshop, Rhinoceros, Keyshot
How can we use AI in the kitchen to reduce food waste?
1 | Identified Problem
I decided to focus in on implementing AI within a fridge and cabinetry space.
I then restructured my problem statement to discovering how this space could reduce waste and improve a user's health and purchasing decisions through pattern tracking and new data collection.
The sketches above are results of my ideation and research phase.
1) The system would be able to understand each unique user's eating and purchasing habits. This would enable better suggestions in relation to these behaviors and how often food goes bad.
2) Suggestions would focus on when a user would need to buy something in the future based on their past behaviors.
3) The user could generate feedback about how foods make them feel throughout the day allowing the system to learn which foods are best for a specific lifestyle.
2 | Visuals
I created this visual to show how data would be represented on the system screen. This was one of my first times using graphics to convey my idea and this process taught me how to effectively translate words into visual design.
3 | Final
The graphics above details the system name Nutrient, that translates to Nutrition to holistically convey the user's experience.
Using artificial intelligence and touch screen technology within both a fridge and cabinetry, this system discovers foods that go to waste most frequently or what foods influence the user's health and mood in order to suggest better purchasing and eating decisions.
With today's abundance of conflicting health data, the Nutrient will form more reliable conclusions for each unique user.