Ripe
Your friendly kitchen navigator.
Role: Research Lead, UI Designer
Brief: Design an interactive prototype based on AI technology
Target Audience: Working millennials ages 25-35 that share common traits of interest in saving time and resources, and eco-conscious.
Applications: Figma, Photoshop, Illustrator, Google
Problem
Upon research, we found that each year, it costs the United States $218 billion dollars of food waste much of which is still perfectly edible, costing the average American approximately $1,800 every year. This is due to various reasons. Aside from rotting from bad weather, processing problems, and overproduction; it mainly comes from consumers’ overbuying, poor planning, and confusion of labels.
Pain Points
Pain Point Cycle
We found that our users lack confidence in the kitchen; whether it’s undereducated tool use or fear of hazards, these problems make meal planning more challenging and time-consuming. It then leads to poor planning meaning food is going to waste, and finally leaves our users feeling defeated from throwing food and money away.
Risks
Market saturation
The Risk here is that many solutions have already been explored around low kitchen waste.
e-excess
People may not want to add another device to their already existing load of devices in their home.
Receptivity
Not everyone may be open to trying for various reasons whether it’s interest or motivation.
Cost
This is always an issue with any electronic device especially with AI implementation because it can get pricy very quickly and users may not be able to afford it.
Eco-motivation
Not everyone is concerned about the environment, hence the low motivation.
Tech-security
This becomes a risk when users feel like they won’t be able to know how to use a product with AI.
Hypotheses
We believe that we can decrease food waste and increase user confidence by creating an inclusive solution that uses machine learning to guide users through their home kitchens.
We will have demonstrated success when users who cook using this tool feel more confident and satisfied after cooking than those who cook without it.
So…How Might We?
Decrease waste and increase confidence in home kitchens using machine learning?
Guide users through the cooking process using kitchen display and voice AI?
Speed up the cooking/planning process?
Help users to explore new foods?
Help users understand what foods they own & how to use them?
Motivate users to come back to a solution and continuously seek to improve their skills?
The initial input is where we will gather all the necessary information about our users before our machine learning algorithm will sort out data and give us the corresponding necessary output according to the user’s initial input. After a tutorial has been selected and used by the user, each feedback from the users will be picked up by our machine learning again for further refinement to best fit the users’ needs.
User Testing
We conducted pre and post-surveys for both rounds of user testing.
5 users for the first sprint and 10 users for the second.
Sprint 1 Methodology
We had two groups of 5; one group without AI assistance, and one with AI assistance. Both groups took a survey before and after the cooking experience, to gauge their preconceptions and experience.
Users without assistance:
Users were tested by cooking on their own. They were given no assistance
Users with assistance:
Users cooked while accompanied by a voice reading the instructions to them throughout the process.
Results
Confidence levels rose for users with AI
Meal Satisfaction was higher in users with AI
Understanding of Ingredients was higher in users with AI
Anxiety decreased in users with AI
Our users didn’t know what “zest” meant.
During our user testing, 6/10 users said they didn’t know the vocabulary used by some of the third-party website’s recipes.
Braise, aerate, zest, char, froth…are these words familiar to you? We’ve heard Gordon Ramsay yell these words out on TV. But what do they really mean?
We made sure our users wouldn’t get lost and guess the process of cooking a meal. Ripe lets users see and read tutorials and guides of each action word between steps.
Mid-fi Designs
Sprint 2 Methodology
Validate the data collection for AI recipe suggestions based on user preferences.
Gather Preferences
Sent out a survey to 27 people and collected information on their personal preferences
Present Recipes:
Filtered and sent recipes to the same survey respondents to gather feedback
223
Recipes sent across 27 cases
77.58% Avg. successes
173 recipes marked as a match
74.04% Case Success
20 cases with >66% of suggestions accepted with all cases having at least 1 acceptance
Final Insights
App Configuration
Ripe has five main modules -- dashboard, schedule, pantry, groceries, and cooking -- all available on mobile devices. The cooking module uniquely integrates with Alexa devices, including both sound-only devices like the Echo and visual devices like the Alexa Kitchen, to guide the user in real-time through the cooking experience.
Ripe's cooking feature walks the user step-by-step through each recipe in real-time.
While you learn to cook from Ripe, Ripe also learns from you.
Its AI and machine learning algorithm do all of the heavy liftings to find recipes and curate custom shopping lists for you.