Essential Pantry addresses food wastage in urban American homes. An average US household of four discards an estimate of $1,350 to $2,275 in annual losses according to Bloom, American Wasteland, pg 187. The reasons around this wastage are many. This includes labeling confusion around when the foods should be discarded, spoilage from suboptimal storage, poor visibility in refrigerators, partially used ingredients, and misjudged food needs. (source Wasted : How America is losing up to 40 percent of its food from farm to fork to land ll).
Essential Pantry allows busy couples and small families, to streamline their grocery buying and consumption by tracking groceries and leftovers and simplifying the decision making around what to make for dinner. The target customer is dual income urban couples with young children. People who are looking to raise a family, eat healthily, eat within budget, and have a career at the same time. The product is a voice-activated hardware device mounted on the refrigerator that allows users to add items they run out of to a shopping list by the touch of a button. These items can then be purchased from grocery partners online for same day or next day delivery. Using the knowledge of what is in the user’s pantry and what they have shopped for, the device can recommend recipes that the user can make for dinner. The product uses machine learning to estimate the items that the user is about to run out of derived from their usage. It also notifies the user around food expiration soon, and whether it should be consumed or discarded.
The Technical Details
We developed this project until the prototype phase, even though we did develop a 'looks like' design, that was just on paper. We developed a functional physical product and cloud-based services. The ‘works like’ prototype included a raspberry pi, a small single board computer, and a raspberry pi shield with an 8x8 LED matrix and a button. The button allowed the user to activate the voice feature and add grocery items that they have run out of, and the LED matrix acted as a visual feedback interface that displays icons indicating whether the action was a success or not. We used an external speaker/microphone at this point for the audio component. The ‘works like’ prototype was housed in a laser-cut acrylic box. The idea was to create a CNC wooden enclosure or a 3D printed enclosure for the prototype. Both these options allowed the LED lights to effectively shine through the material. We also developed all the UI for the mobile and the web platform and had it working using google polymer.
Concept Design, Interaction Design, Physical Design, Technology Development, User Testing
June 2016 - June 2017